Continuous Blending with homemade stik

[Conch]

diverb,

If you are making the one in the ox hackers guide here are a few things we found. Just made two of them. The main body is ok at 2" diameter, but if you drop it down to 1" it will put a load on some larger compressors. We tested it and the vaccuum gauge went through the roof on a 18 CFM Mako. Then we took out some of the mixer baffles .... only a couple are needed ... don't drill the 4 holes in them, cut the tops off and leave the baffle plates there. I found that you can get a piece of PVC that will fit right in and slot it in a few spots for the baffles. this slides in easier than the caps that the hacker guide recommended. This way you can control the amount of restriction. We ended up making the bottom end 1 1/2" and with 3 baffles it worked fine. I would also put a larger filter on the top for better air flow.

The one that is in the hacker manual will probably work fine on smaller compressors. It didn't on the 18 CFM we were testing it on.

Also, be careful with the regulator. A metering device should be used between the reg and the stick intake. This way no one can accidently crank up the o2 and blow up the compressor. I'm still looking for the "right" one for mine

Conch

 

[oxyhacker]

18 cfm is a lot of compressor for that little tube! The book says that the one shown is good for up to 8 cfm or so, and suggests scaling up the design for larger compressors - going a size or two larger on all the pipe/fitting sizes would probably have solved all your problems. It's also important to adjust the design to the compressor it will be used on by making sure the passages at the most constricted points (going through the baffles) are roughly equal to the size of the compressor inlet tract (which you did by chopping the tops off!).

I'm not surprised you don't need much in the way of baffles - with 18 cfm going through that 2" tube there must be a heck of a lot of turbulence! Using a pipe inside to hold the baffles works fine, only thing is it further reduces the ID of the pipe, which could run up the resistance unecessarily high with a higher output compressor; I like the drilled caps because they combine they increase the gas velocity and turbulence just as they hit the plate which folds the flow over onto itself, but lets face it, just about anything that messes up the flow will work.

The restrictor thing is pretty easy if you get a good quality Swagelok or Parker locking instrumentation needle valve and set it the way the Nitrox Stick does, as described in the book. The sleeve on the reg adjusting screw works OK too, and makes a good stopgap until one can find the right valve. Actually, using a sleeve on the regulator adjustment knob, as well as the needle valve works particularly well, as it can bring down the adjustment range to make fuller use of the tank contents.

Originally posted by Conch
diverb,

If you are making the one in the ox hackers guide here are a few things we found. Just made two of them. The main body is ok at 2" diameter, but if you drop it down to 1" it will put a load on some larger compressors. We tested it and the vaccuum gauge went through the roof on a 18 CFM Mako. Then we took out some of the mixer baffles .... only a couple are needed ... don't drill the 4 holes in them, cut the tops off and leave the baffle plates there. I found that you can get a piece of PVC that will fit right in and slot it in a few spots for the baffles. this slides in easier than the caps that the hacker guide recommended. This way you can control the amount of restriction. We ended up making the bottom end 1 1/2" and with 3 baffles it worked fine. I would also put a larger filter on the top for better air flow.

The one that is in the hacker manual will probably work fine on smaller compressors. It didn't on the 18 CFM we were testing it on.

Also, be careful with the regulator. A metering device should be used between the reg and the stick intake. This way no one can accidently crank up the o2 and blow up the compressor. I'm still looking for the "right" one for mine

Conch

 

[FredT]

A sphere packed tube is also a good mixer.

4 to 6" PVC and ping pong ball sized media is one other choice when doing higher flow rates. Cooling tower fill modules (hollow high surface area molded bulk filler) is another good mixing medium if you have them available. Cooling tower fill usually comes in _large_ volume multiples though.

Inexpensive plastic spheres can be had from United States Plastic

http://www.usplastic.com

Part numbers 2101 through 2108. Sphere diameter should be 1/3 tube ID or smaller. Assume flow area to be 1/4 of unfilled tube cross section for single size balls. Use the compressor inlet port diameter as the minimum cross section reference.

As an example if the compressor input is 1.5" pipe diameter, that works out to 1.8 square inch flow area. Remember pipe is measured by INSIDE diameter! The pipe OD in this example 1.900". To get 4x that flow area you'd need 3" pipe. The smallest ball available is 3/4", so that meets the 1/3 diameter requirement but the next size up 1.5" ball is too big. Allow at least enough length to get 10x the diameter of your ball stack. 20X is better if you have the balls to fill it. Also be sure your end screens have at least 2x the flow area of your input with maximum hole size about 1/4 the diameter of the sphere fill.

Small spheres come in cases of 1000, or 500 for the larger sizes, so even if two to four people split a case, all will have more than enough to make a 10x mixer. PVC pipe comes in 10' or 20' lengths. Do a test fill and trim the pipe length to suit the fill available.

A sphere packed mixer does not _depend_ on turbulence for mixing, although a bit of turbulence is a good thing where mixing is desired. Even at low flow rate the shed eddies and flow dividing provided by a serial ball stack will do a good job of mixing.

FT

 

[oxyhacker]

As long as the spheres aren't small enough to be sucked down the intake

Seriously, I saw one which used marbles. Actually worked really nice, but if a little piece of screen had crumpled, it would all have all been sucked down. Still, as you say, for a high volume mixer it's a neat idea.

Originally posted by FredT
A sphere packed tube is also a good mixer.

 

[FredT]

Originally posted by oxyhacker
As long as the spheres aren't small enough to be sucked down the intake

With large diameter balls that is less of a problem. I envisioned a PVC or SST plate with a grid of 1/4" holes drilled in it as the downstream "screen". Metal mesh on the downstream end would be asking for a compressor rebuild, although the ball would not make it past the reed valves in the first stage intake head of most 3 or 4 stage compressors. The other safety factor would be to mount the stick vertical with input at the bottom.


BTW it's good to see you here Vance. It's been a while since we talked regularly on rec.scuba.

FT

 

[The_DivePirate]

What would you think about golf balls? 4"x36 piece of pvc, 20 or so golf balls, a shower drain grate, and so on? Should this unit supply a 15cfm?

 

[fire_diver]

Even though this thread is ancient, I'll thow my .02 cents worth in here. I don't believe the golfball idea would work suitabley. IIRC the air needs to be "folded" many times in the stick. The balls just make the air a little bit turbulant.

Comrade Stroke

 

[amascuba]

Found it.

Patent no. 05992464

or try this URL if your browser can fit it all:

http://patimg2.uspto.gov/.piw?docid...=6%26f=G%26l=50%26d=ft00%26S1=nitrox%26OS=nit

Your search expired. I found it on google's patent search.

Pre-Compression Nitrox Inline Blender: Patent 05992464

 

[waterpirate]

I just went down this road and here is what i ended up with. 4" pvc pipe for chamber, filled with layers of golf practice balls" the wiffles". seperated into chambers with pvc drain grates, some with slots, some with holes. Talk about your tortured path!! I used a 4x2 reducer on the top to fit my chrome holley carb filter. drilled and tapped it just under filter for o2 port. Bottom 4x11/2 reducer to match bauer intake, drilled and tapped for el-cheapo sensor. 02 is controlled using welding regulator with brass sleave to prevent exceeding 40%.
works great, less filling!
Eric

 

[The_DivePirate]

That's what I'm talking about, but I may have to go to old faithful and just upsize it.