Needle valve tiny adjustment = huge bubble change

Here's a couple of pictures of the inside of my stand.

What you can't see from this picture is that the CO2 tubing goes outside my stand on the left into the glass inline bubble counter, through another brass valve and check valve and then re-enters the picture at the reactor.

I think I over-engineered my system and in trying to make it too serviceable and too fail-safe I induced too many places for leaks. I thought the idea of the brass valve would make it easier to re-fill the bubble counter without leaking water from the injection point out of the reactor. I just close that valve, unhook one side, refill counter with water, re-hook back up and away we go...no water leaking from the reactor. And now I read that these brass valves are only rated for 10 psi...oops.

Brass Valve- Quality brass valve for years of reliable use. Thumb screw valve for precision control. 10-psi max.

This weekend I can try and run a line with as few connections as possible...regulator to bubble counter to reactor. I'll see if that makes any difference. But that was essentially how I had it set up on my previous 10 gallon with a diffuser instead of reactor. I still had the same problem on the last tank, maybe a bit less pronounced.

Now I'm thinking maybe a metering valve problem. Maybe I'll take that apart this weekend first, before trying to replumb everything.

More detailed view inside stand:


And older picture of stand innards when first set up. Regulator not connected yet, was just doing a DSM at the time.

Oh, that's pretty

I have had valves like such where the finest of movements would extremely alter bubble count. What a pain, especially when your only trying to up the bubble count by 1, LOL! Sometimes you'll find me just tapping the needle valve to get that little bit of movement.

Things I've done to gain more control is place a second needle valve in line. Replace soft CO2 tubing with ridged tubing (this minimizes expansion between regulator and check valve, and between check valve and defuser).

If the check valve is spring loaded, I suggest replacing with a diaphragm type.

What could be happening here is there is back pressure created either by the check valve or the pressure of the diffuser. If this is the case any adjustments you make at the needle will mess with your head. You'll make an adjustment and it may look okay for a minute. You come back and CO2 could be to high because it took time to build pressure, stretch the soft lines, then when the check valve finally opened up more, blasted CO2 into the defuser...

This is the 1st time I tried explaining this so I hope it made since.

To recap, possibly change the tubing to ridged hard tubing. Make sure the check valve is a diaphragm type and not spring loaded. If that doesn't help maybe just put another needle valve inline.

before you take the metering valve apart, please put a bubble counter right at the outlet of the metering valve, to see if the bubble rate is manageable.

oldpunk78 said:

I just reread your 1st post. I was about to ask if this was how it's always been or if it was something new. I've never owned that particular metering valve but i suspect that it just sucks. If the problem was new i would be suspicious of a broken needle or a leak as betatail suggested. Also, you are miss informed as to the fact that you have a s_series. It is not. It's an older higher flow valve.

are you able to go from almost no bubbles to a lot? If the needle is broken, it wouldn't do much in terms of predictable adjustment. Habe you checked for leaks?

Click to expand...

it was new....
I built it, but anything after the metering valve, not my creation.
was it a gift from a friend or did you buy it from some one in Seattle?

here is the picture..

Heard someone mentioned my name . I have the same valve (inline version). Im running about 3 bubbles per second on my 10 gallons. As you can see in the picture. It takes about 4 full turns to get to 3 bps. This is just my speculation. You are running a reactor so there isn't back pressure as you would with a ceramic diffuser. I tested my s series with a bubble counter that was not connected to a diffuser (pix below), and it took less turn to increase the bubbles. By the way there isn't any valve that sucks (exclude non industrial). If you can get down to 1-2 bubbles per second then you're good.

Yep, I bought this from you Bettatail (I have a different username on Barrreport, why a different name, I don't know?!). It worked well on my previous little 10 gallon connected to a diffuser. That was a bit of a simpler setup going from regulator to inline check to bubble counter to ceramic (non-atomizing) diffuser. I was surprised to find I had to get up to 40 or so psi to get adequate control but once that was set I rarely changed anything on that tank. I still required some finesse in changing the bubble rates on that tank, just little 1/16th or so turns would make a noticeable difference. But it was definitely manageable on that tank. This current set up is quite a different animal. I'll do some serious leak checking tonight, I've done it in the past with soapy water and didn't find anything but maybe I didn't have enough flow to make noticeable soap bubbles.

I'll try and make it a much simpler system this weekend and see where it all ends up.

I can get less than 1 bps and definitely upwards so the valve may be fine. Let's hope it is just leaks.

Thanks for the help everyone!
**I updated the original post to remove the s-series. That was my mistake trying to get too specific on the type of valve before I went and grabbed the model # off the side of it.

Late for work :|

Ya'll be shocked when you hear what I found.....a leak....shocking right?!

I'm ashamed I didn't find it before because it was quite large. It was at a Clippard Quick-Connect which only has a single 1/8" barb. I recently changed out the tubing around this fitting and used whatever I had sitting close by. In this case it was rather thin-walled air hose. It should have been pretty obvious that this wasn't going to fit the bill because it was kind of loose but I thought the single barb would hold on OK. I was wrong. Probably when I went up in pressure to 20 psig it didn't retain a decent hold and began to leak.

The soapy water I initially used last week to check for leaks wasn't very sudsy either. I used some fancy-environmentally friendly-biodegradable-hypoallergenic-too expensive dishsoap that really doesn't bubble well at all. I bought some leak checking solution at home depot (yes, this is just $5 soap in bottle but it had a fancy dobber!) and it led me right to this leak. Lesson learned. Use normal dishsoap or this bottle of leak checking solution that I'll probably lose well before I use it all up.

I replaced this tubing with some thicker walled 6mm OD/4mm ID tubing and placed a tiny hose clamp (from auto parts store) to help keep it secure.

Working on dialing it in right now. Starting out on the lower end of pressure, 15 psig, a two number increase on the vernier handle (see AlanLe's first picture for what I'm calling a two number move...for instance I moved it from the number 10 to number 12 as visible on his picture) moved me from 2.2 bps to 3.3 bps. This is a marked improvement from what it was before. I'm in between 2 and 3 full turns open. No problem moving it down to less than 1 bps so I'm concluding the valve is just fine.

I removed one of the brass stop valves and shorted the tubing quite a bit to provide faster response time to my changes and to reduce the pressure drop. I might possibly be able to go a little bit lower in pressure, but not much. The swagelok check valve needs 5 psig (ss-4c-5) so that's definitely the lower limit.

Am I correct in assuming a lower pressure will allow finer control of the flow (i.e. more turns to see a change in flow)? Or do these valves work a little better when they have some pressure behind them?

Thanks for everyone's help. Sorry it wasn't some earth shattering problem. Bettatail built a solid regulator with nice components. This was totally a problem I caused.