Some thoughts...

I have been doing some thinking on why the R-600, R-170, R-14 charge did not work as expected. It appears to all come down to how effective is the separation in my home brew phase separators. I had assumed it would be simply adequate, but not phenomenal, especially in the later stages. But now I'm not too sure about this.

Remember the 3600 Watt "Big Brother" unit I showed in a previous photo? A unit similar to this was my first test bed for an R-600 charge. It worked unbelievably better, and with only half the liquid (R-600) than the previous R123/R22 combo. So I just assumed that I would see similar results with my AC-2 unit when using R-600. Why didn't I? well I think that proportionally speaking, all my phase separators are much much larger then what is used on the "Big Brother" unit. How much larger? At least 13 times the volume (based on compressor mass flow differences).

So it would stand to reason that the AC-2 unit would be far superior in its ability to separate the liquids from the gases at each phase separation point. Having poor phase separation allowed the "Big Brother" unit to get by without a mid-carrier refrigerant between R-600 and R-170 (80 degree C difference in boiling points).

But the AC-2 unit with its superior phase separation, ended up disconnecting between the lower and upper stages, mainly due to excessive amounts of R-170 being dissolved into the R-600 and robbed from the downstream stages. And later as the lower stage got even colder due to it's isolation from upper stage loads, the temperature dropped to the level that even R-14 was being gobbled up as well. Probably the best way to over come this would have been to add a mid-carrier refrigerant such as R-290 or R-22. Another way was by using brute force, and adding excessive amounts of R-170 and R-14 in order to compensate (this is what I did on my first tests). But from a thermodynamic aspect, this probably just ends up making 3 phase separated autocascade, into something that acts more like 2 phase separated design due to inter-stage flooding.

R-600 should still be better at moving heat then an HCFC. So perhaps I'll try a future test with an R-600/R-22 combination to see what happens. If I had R-290 I'd try that as well.

I still might get even more out of the AC-2 unit