Thoughts on filling ones own tanks

[pensacoladiver]

I think I follow. But you assumed one critical piece of information.... that I am smart enough to figure out the volume in CF.

My first filter is 8.5 inches tall and 1.75 inches radius. According to my formula, that gives me 82 cubic inches or .048 cubic feet.

2nd filter is 10 inches high with raduis of 1.5 which gives me 70 cubic inches or .041 cubic feet.

Next step gives me 8.16 cuft at pressure in first filter and
6.97 in the 2nd filter.

Does that sound about right?

[CaptnJack]

15 cf is about right. My smaller chambers are functionally about 12cf.
So assuming you are filling hp100s stop pumping He and start pumping 32% about 250psi short of what your helium target pressure. (hp100s have a tank factor of 6, that is 6cf per 100psi)

[Jeff Pack]

Sorry to revive an old thread, but I'm just gettign ready to soon start mixing trimix.

Whats a good regulator to start with on the helium side? I'll be doing continuous blending, although I'm abit confused above on using only 1 mixing stick.

I thought somewhere, but now I have to find it, was to use 2 mixing sticks in series, I think introduce helium first read the output, then introduce O2 in the downstream stick. read the output at that point. But now I have to go find all that again.

[kdupreez]

Sorry to revive an old thread, but I'm just gettign ready to soon start mixing trimix.

Whats a good regulator to start with on the helium side? I'll be doing continuous blending, although I'm abit confused above on using only 1 mixing stick.

I thought somewhere, but now I have to find it, was to use 2 mixing sticks in series, I think introduce helium first read the output, then introduce O2 in the downstream stick. read the output at that point. But now I have to go find all that again.

One stick and one O2 sensor works fine.. you have to make sure you have a decent stick that can create enough turbulence to mix before it reach the O2 sensor..

- Start Compressor and calibrate O2 sensor for 21%
- Dial Helium till Oxygen = (1 - Helium Fraction of desired mix) x 21
- Dial Oxygen till it reads your desired mix..

example: if you want 21/35

1. Calibrate to 21%
2. Introduce helium untill O2 reads 13.7%
3. Introduce Oxygen till O2 reads 21%

depending on your compressor and blow by, you might see this drift a little and helium get higher and O2 lower.. the first couple of times adjust helium to reset O2 levels back to 21.. then adust Oxygen after that..

You can also measure your final outcome as it fills the tanks.. I just hold the tube into the din bleeder and analyze, then slightly adjust accordingly..

I've been usually good with a +/- 0.5% variation in mixes..

IMPORTANT: every 1% that you are off on the initial calibration could be about +/- 4% on the final helium!! so you have to be very precise with that initial calibration and helium addition..

[pensacoladiver]

I used koos's method once and I ended off just a bit on the final mix. I think it's because I have oxygen creep as the tanks fill.

As the oxygen started to creep, it was really a guessing game as to whether to increase helium or decrease oxygen. I did a little of both and ended up off

Can I ask why you are set on continuous mixing?

I use Richards method of pumping straight helium through the compressor and then topping with nitrox.

Works like a champ and I usually nail the mix within 1 percent on the helium and dead nuts on the oxygen.

Another benefit is that you are not pumping helium through your compressor at high pressures when you do it this way.

Of course, if you are using lp tanks, it's pretty much a moot point.

[pensacoladiver]

First pic is my helium bag hooked up to the compressor. As long as you keep the bag somewhat pressurized all is good.

Second pic is my needle valve that I use to keep helium in the bag. It's from McMaster Carr and holds a constant pressure very well. Some of the cheaper valves require constant adjustment to keep the bag at a good pressure.

[pensacoladiver]

I forgot to add, in the first pic you can see my normal input tube coming down from my nitrox stick. When I'm done with the helium I just pop one grey tube out, stick the nitrox tube back in and am filling with nitrox within about 15 seconds from the time I stop pumping helium.

[kdupreez]

Thats pretty cool.. I've seen richard's and its a very elegant way to pump other inert gasses through the compressor.

I've also seen a similar type thing where a scuba first stage connected to the helium tank and second stage regulator with exhaust valve plugged shut and the mouthpiece was affixed to the compressor inlet and then sucks "on demand" gas out of the helium tank. no need to adjust any needle valves, but potentially fairly expensive compared to a trash bag.

[pensacoladiver]

I thought about going with the regulator setup but that puts a lot of cycling on the regulators and I have read it might be potentially too hard on the regs and seats.

You are right, a zip lock bag and 2 spot ties is a pretty cheap way to go for trimix. :-)

[Jeff Pack]

If I didnt have o2 creep, I'd use the one stick whirlygig, but the problem is my o2 doesnt stay fixed, I have to slowly decrease the feed of o2 when mixing Nitrox, so in a single stick, its kentucky windage.

Whereby with 2 sticks, 2 sensors,theres no question (or shouldnt be anyways) a question of what needs to change.

But again, I'm just going by memory when I was first reading up on all this. I've got a spreadsheet I pilfered from somewhere that shows the values whether you mix he first stick/ O2 second, or the inverse. As I recall, its just math which way it goes.

But still could use some help on a helium regulator though.

[pensacoladiver]

When I did my one run with cb I used a Chicago welding reg available all over eBay. I have 2 because I like to drain my oxygen tanks down to nothing before refilling. 2 regs allow me to do this.

A plain old oxygen flow meter will give you precise control over the helium I'm not concerned about the Lpm flow rate as there is still plenty if room to open it up once the ball reaches maximum listed flow on the guage. And if you are going to be using multiple analyzers, the flow guage doesn't matter anyways.

I set regulated pressure (2nd) guage to 40 psi.

If you defied to go with the ziplock bag method, you don't need a regulator, just a needle valve.

[Jeff Pack]

thx!

[Jeff Kruse]

Please, if you have a compressor and your diving trimix you can afford a He analyzer. Then it’s just too easy to CB trimix. CBing is fantastic. It lets you get that last bit of He in every bottle.

[Jeff Kruse]

Please, if you have a compressor and your diving trimix you can afford a He analyzer. Then it’s just too easy to CB trimix. CBing is fantastic. It lets you get that last bit of He in every bottle.

[kdupreez]

He analyzer is not the problem.

The issue is he analyzers require flow through the analyzer its much harder to get one inline on he intake.

You can't just stick a sensor into the stick and be done with it.

Hence putting he analyzer on the the fill whip and measure output.

But the output is about 2.5min behind any intake adjustments due to dwell time in filter stack.

With slower compressors that could be 5-6min.


So applying he sensors at the intake is hard unless you have a dedicated tmx analyzer for that or some small vacuum pump to circulate through analyzer.

[pensacoladiver]

I have a helium analyzer but as koos says, it's not hooked into the system.

Filling through the bag into the compressor does allow all helium to be used.

I even hook a second tank and whip into the bag once the first tank starts to get low so I can truly use ALL gas out of the first cylinder.

Like I said before, I don't like pumping he through my compressor at high pressures and cm forces me to do that, even if it is only a 35 to 50 percent mix.

[Jeff Kruse]

Yes there is a delay but you anticipate that at the beginning and end of a fill. It’s easy.

Also, you should be using flow gauges for the He and O2. The flow gauges have needle valves built in. Makes it a snap to dial in your mix. The flow gauge gets attached to a regulator.

Yes, you need to have a minimal flow from the output of the compressor going into the analyzer. I assume your fill panel has a manifold. Just put a cheap reg on it and a flow restrictor on your sample line and your done.

Once you get the hang of it its easy. I would start filling the tank when I turned the compressor on. I would turn on the He and O2. A few minutes later I would see the mix at the output. Depending on how many tanks I needed to fill I would adjust the mix. The 3+ minutes of air in the beginning AND end would require me to pump 21/40 but that way I wouldn’t waste any He. The He would be within a few percent.

[camerone]

Sensidine, Parker, and others make miniature diaphragm pumps which show up on eBay all the time. If you really want to CB, then it's pretty easy to plumb in a sample tube at the end of your mixing chamber, using the small diaphragm pump to pull a small bit of the gas mix out and feed it under flow to the helium sensor. You can return the output back to the stack if you want so nothing is wasted, although to be honest, the sample rate is so low that it's immaterial.

Alternatively, as has been suggested, two mixing sticks in series along with a pair of galvanic O2 sensors and a little math relieves you of the complicated plumbing and pumping needed to directly measure helium. (Actually, it's an indirect measurement no matter how you do it - either speed of sound like DiveSoft, or wheatstone bridge against a known sample like everyone else...but I digress).

The geek in me would probably just program up an Arduino to do the math and run a small digital display for everything... of course, I've got some instrument grade ADCs and OpAmps lying around in the junk box