As long as your gasses are pure+/- then measure any 2 and you know the 3rd. The assumption's are that the gasses are pure enough to do this, and you compensation for the inputities are done correctly. You can do this also by measuring only the O2. Some assumptions are safe to make others are not. for instance if i know i used good gasses and i have a trimix gas made of helium and AIR, then a O2 reading of 10% makes the mix 10/50. that will be off a bit cause the air is not O2 and N2 exclusively. There is that pesky 1% of argon in the air portion.This same issue has always been a factor when using concentrator's when making nitrox. All you can get is 96% O2 from one of them. How much error depends on how much of concentrator O2 is used. Then 20+% gas has error and <20% gas has little if any. Non concentrator O2 has for all purposes 0 effective error. Measuring O2 and He is a good measuring method. Measuring O2 and n2 is now being done on assumptions. Im not an expert by any means but i would look at is as measuring O2 and measuring inert gas. If your O2 reading is correct, (assumption safe) and your measure the n2 (assumption safe) then the estimated He will have some error involved because of other gasses not usually accounted for. so a 25/25 gas may be 25 O2 24 He 50 N2 and 1 other if concentrator O2 is used. I cant say that i have ever seen a shop use Nitrox from a concentrator to blend trimix. Home blenders would probably have this issue. The question becomes: So you compensate for the variables when making assumptions in the right directoin to make the estimates He remaining an accurate enough figure. I cant tell ou how many times i have had people tell me that my concentrator O2 is 4% short on O2 from being pure, and therefor my O2 meter has to have a 4 Percent adjustment made to it. Most of us know that is crap, but those that think the logic is solid is now not diving with the mix they think they are.
