• You are viewing the forum as a Guest, please login (you can use your Facebook, Twitter, Google or Microsoft account to login) or register using this link: Log in or Sign Up

Guide to choosing the right needle valve for precise control of CO2

Marcel G

Member
Joined
6 Dec 2024
Messages
75
Location
Czech Republic
Acknowledgements: In writing this article I was inspired by a post by @Bettatail on the PlantedTank.net forum, to whom I would like to thank.

Calculating flow rate

1737870837346.png

What valve (flow rate) do I need?

At flow rate of 30 bubbles (Ø 4 mm) per minute, you add 1 mℓ of CO2 gas per minute [to your aquarium], i.e. 60 mℓ (= 0.06 ℓ) per hour.

Volume of 1 bubble (BV) = ⅙ * π * d³ = ⅙ * 3.14 * 43 = 0.166 * 3.1416 * 64 = 33.4 mm³
d = diameter of 1 bubble (in mm) ← you have to measure or estimate this [using a photo where you capture the bubble near the ruler]
Volume of 30 bubbles = BV * 30 = 33.4 * 30 = 1,002 mm3 = 1 mℓ

If you want to increase the amount of CO2 [say] from 1 mℓ to 2 mℓ per minute (i.e. from 30 bubbles to 60), then your needle valve must be able to increase the amount of flowing gas by just 1 mℓ per minute [using the smallest possible turn of the control knob (valve handle)].

Assuming that the smallest possible turn of the control knob is 1/20 of a full turn (which is a really tiny turn), you will add about 20 mℓ of gas to the aquarium after one complete turn of the control knob (360°). If the maximum number of turns of the needle valve is [say] 20, then if you open it fully [from 0 to 20, i.e. from closed to fully open], you will achieve a flow of about 400 mℓ (= 0.4 ℓ) of gas per minute.

Thus, if you want to be able to regulate the gas flow smoothly, in increments of say 1 mℓ per minute (i.e. 30 bubbles per minute = ½ bubble per second), then you need a valve with a maximum flow of ~0.5 ℓ/min.

If you have the outlet pressure on the CO2 cylinder (= pressure after the pressure reducing valve) set to 45 psi (3.1 bar), then you need a needle valve with a flow coefficient (Cv) of approximately 0.001. At an outlet pressure of 20 psi (1.4 bar) you need a valve with a flow coefficient of about 0.0024.

Model graph of gas flow vs. valve handle turns (for Hoke Milli-mite 1300 series valve):

At an inlet pressure of 45 psi:

1737872197560.png
As you can see (in the above picture for 45 psi), each full turn of the handle of this valve represents an increase in flow of about ~0.6 ℓ/min (except for the very first turn). A 1/20 turn (18°) [which corresponds to one part on the vernier scale] represents a change in flow of ~0.03 ℓ/min (= 3 mℓ/min). So turning the handle one tick (part) on the vernier scale changes the flow rate by 3 mℓ/min, which corresponds to 90 bubbles per minute (= 1.5 b/s).

Hoke Milli-mite 1335 needle valve with black vernier scale handle with twenty ticks (marks) ↓

1737872235129.png

Legend

Flow (ℓ/min):

1737871468396.png

Basic terms
  • Orifice size: the smaller the better (most valves have an orifice of around 1 mm)
  • Flow coefficient (Cv):the smaller the better
    • flow coefficient of most valves is linear increase proportional to the number of turns to full open
  • Flow (Q):ideally below 1 ℓ/min [at full open] (or at least in units of litres per minute)
    • Note: All flow rate data for individual valves corresponds to an inlet pressure of 45 psi.
  • Turns to full open: the more the better (most valves have 10-20)
Comparison chart

1737871965169.png

Tip

A number of excellent and very good valves can be found on eBay for a decent price. [P.S. Although sometimes vendors offer a valve at an exorbitant price, I recommend bargaining with them offering a lower price (e.g. $50-90) for the valve. Some will accept it without any problem.] The most common valves you can find there are:
  • Hoke Micromite 1600 series
  • Hoke Milli-mite 1300 series
  • Whitey 21/22 series
  • Swagelok S series
New valves of these brands [and models] are also relatively affordable:
  • Aalborg MFV (VM1/VM2)
  • Ideal valve 52/54-1
  • Ham-Let HXF 1300 series (1°)
To be continued ...
 
Excellent & very good (low flow) metering valves

There are certainly a number of fine needle valves available on eBay (and elsewhere) that are not listed here. The reason is that I have been unable to find technical documentation for them, or at least data on their orifice size (or flow rate) and number of turns to full open. If anyone knows of any other good (maybe even older, now officially no longer sold) needle valves and would be willing to share it, he can post it here.

Swagelok

S series (1.9 ℓ/min)

1737874725415.png
Swagelok SS-SS2-VH
  • Web: swagelok.com
  • Part number: SS(B)-SS*, SS(B)-SM*
  • Orifice: 0.032" (0.8 mm)
  • Cv: 0.004
  • Turns to full open: 10
  • Price:
    • official: $400-700?
    • ebay: usually $150-200
  • PDF: swagelok---s-m-31.pdf
  • Thread: male/female NPT, Swagelok tube fittings, male VCR fittings
    • VCR is a special connector to which the hose/tubing can be attached (so you don't have to buy any additional connector)
  • Shut-off capability: no
    • The handle stop is set at 4-10 mℓ/min with 15 psi inlet pressure. Adjusting stop to lower flow setting can damage valve and stem tip!
Graph of gas flow vs. valve handle turns (for Swagelok S series valve) at an inlet pressure of 20 & 45 psi:

1737874815622.png

Flow (at 20 psi):

1737874865920.png
  • mℓ/min = sccm (cm3/min)
  • 1⁄20 = 1⁄20 of the 1st turn (the smallest possible fraction of one turn)
  • bps = bubbles per second (Ø 4 mm)
  • bpm = bubbles per minute
Swagelok (Whitey)

21/22 series (3 ℓ/min)

1737875011479.png
Whitey B-22RF4


  • Web: discontinued
  • Part number: -21???-, -22???-
  • Orifice: 0.02" (0.5 mm)
    • there are two versions (21 vs 22?), different by orifice size, the large orifice size older version is not accurate at low flow
  • Cv: 0.007
  • Turns to full open: 8
  • Note: These are the mystery discontinued Whitey precision low flow control valves.
  • Price:
    • official: not sold anymore
    • ebay: usually $150-200 (under $65 if you're lucky)
  • PDF: unavailable
  • Thread: male/female NPT (?) or UNF (gyrolok)
  • Shut-off capability: unknown
Graph missing …
Flow table missing …
  • Flow (at full open): ~3.3 ℓ/min
To be continued ...
 
Last edited:
Hoke

Milli-mite 1300 series (4.7 to 11 ℓ/min)

1737875454884.png
Hoke 1335M4B


  • Web: hoke.com, circlevalve.com
  • Part number:
    • 1° stem: 1335*, 1345*
    • 3° stem: 1315*, 1325*
  • Orifice: 0.047" (1.2 mm)
  • Cv:
    • 1° stem = 0.010
    • 3° stem = 0.024
  • Turns to full open: 18
  • Price:
    • official: $340 (?)
    • ebay: usually $150+ (under $75 if you're lucky)
  • PDF: hoke---metering-valves.pdf
  • Thread: UNF (gyrolok), NPT
    • gyrolok is a special connector to which the hose/tubing can be attached (so you don't have to buy any additional connector)
  • Shut-off capability: unknown
Graph of gas flow vs. valve handle turns (for Hoke Milli-mite 1300 series valve) at an inlet pressure of 20 & 45 psi:

1737875651261.png
Flow (at 20 psi):

1737875677483.png
  • mℓ/min = sccm (cm3/min)
  • 1⁄20 = 1⁄20 of the 1st turn (the smallest possible fraction of one turn)
  • bps = bubbles per second (Ø 4 mm)
  • bpm = bubbles per minute
Micromite 1600 series (0.4 ℓ/min)

1737875771767.png
↑ Hoke 1654002 (1/4" OD to 1/4" UNF)

  • Web: hoke.com
  • Part number: 1654*, 1656*, 1666*
  • Orifice: 0.031" (0.79 mm)
  • Cv: 0.0008
  • Turns to full open: 18
  • Price:
    • official: $600-700 (?)
    • ebay: usually $300+ (under $100 if you're lucky)
  • PDF: hoke---metering-valves.pdf
  • Thread: UNF (gyrolok), NPT
    • gyrolok is a special connector to which the hose/tubing can be attached (so you don't have to buy any additional connector)
    • 1737875943644.png
    • ↑ Hoke 1654002 (6 mm OD to 1/4" UNF)
  • Shut-off capability: no
Graph of gas flow vs. valve handle turns (for Hoke Micromite 1600 series valve) at an inlet pressure of 20 & 45 psi:

1737875984283.png

Flow (at 20 psi):

1737876010707.png
  • mℓ/min = sccm (cm3/min)
  • 1⁄20 = 1⁄20 of the 1st turn (the smallest possible fraction of one turn)
  • bps = bubbles per second (Ø 4 mm)
  • bpm = bubbles per minute
To be continued ...
 
Last edited:
Parker Hannifin

HR series (2 to 10 ℓ/min)

1737876320750.png
Parker Hannifin N & HR series

  • Web: parker.com
  • Part number:
    • type 3: -H3?--- (orifice: 0.44 mm, Cv: 0.021)
    • type 2: -H2?--- (orifice: 0.37 mm, Cv: 0.014)
    • type 1: -H1?--- (orifice: 0.26 mm, Cv: 0.007)
    • type 0: -H0?--- (orifice: 0.04 mm, Cv: 0.004)
    • Note: There is a discrepancy in the technical documentation between the data on page 8 (Cv for H0 = 0.00034) and its graphical representation on page 11 (Cv for HR0 = 0.004). In response to an enquiry to the manufacturer, I was told that the correct Cv value for H0 is 0.00034 (so the graphs in the documentation do not reflect reality).
  • Flow (at full open): ~1.9, ~3.3, ~6.6 and ~10 ℓ/min
  • Turns to full open: 15 (±1)
  • Price:
    • official: ?
    • ebay: ?
  • PDF: parker---n+hr-series.pdf
  • Thread: NPT?(some models with a-lok connections)
    • a-lok is a special connector to which the hose/tubing can be attached (so you don't have to buy any additional connector)
  • Shut-off capability: yes
    • Most valves cannot be completely closed so that absolutely no gas flows through them. This valve allows complete shut-off = zero flow.
Graph of gas flow vs. valve handle turns (for Parker Hannifin HR series valve) at an inlet pressure of 20 & 45 psi:

1737876493121.png

Flow (at 20 psi):

1737876521771.png
  • mℓ/min = sccm (cm3/min)
  • 1⁄20 = 1⁄20 of the 1st turn (the smallest possible fraction of one turn)
  • bps = bubbles per second (Ø 4 mm)
  • bpm = bubbles per minute
Ham-Let

HXF 1300 series (1° stem) (2 ℓ/min)


1737876573039.png
Ham-Let HXF1380-SS-GL-1/4-S-EP-V

  • Web: ham-let.com
  • Part number: HXF13??--
  • Orifice: 0.03" (0.8 mm)
  • Cv: 0.004
  • Turns to full open: 11
  • Price:
    • official: $160-230 (?)
    • ebay: rare item (under $50 if you're lucky)
  • PDF: ham-let---h1300.pdf
  • Thread: NPT?(most models with let-lok connections)
    • let-lok is a special connector to which the hose/tubing can be attached (so you don't have to buy any additional connector)
  • Shut-off capability: no
Graph of gas flow vs. valve handle turns (for Ham-Let HXF 1300 series valve) at an inlet pressure of 20 & 45 psi:

1737876676228.png

Flow (at 20 psi):

1737876699183.png
  • mℓ/min = sccm (cm3/min)
  • 1⁄20 = 1⁄20 of the 1st turn (the smallest possible fraction of one turn)
  • bps = bubbles per second (Ø 4 mm)
  • bpm = bubbles per minute
To be continued ...
 
Last edited:
Fujikin

PUN/UN/DUN series (7 to 2 ℓ/min)


  • Web: fujikin.co.jp
  • Part number:
    • UN-14M*, UN-34M*, UN-94M*, UN-04M*
      • Cv = 0.015 or 0.03 (at 14 turn full open)
    • DUN-32PGA
      • Cv = 0.015 (at 11 turn full open)
    • PUN-913P-6.35
      • Cv = 0.004 (at 11 turn full open)
  • Flow (at full open): ~1.9 and ~7.1 ℓ/min
  • Price:
    • official: ?
    • ebay: ?
  • PDF: fujikin---pun-un-dun(japanese).pdf
  • Thread: ?
  • Shut-off capability: unknown
Graph of gas flow vs. valve handle turns (for Fujikin UN series valve) at an inlet pressure of 20 & 45 psi:

In progress …

1737877019650.png

Graph of gas flow vs. valve handle turns (for Fujikin DUN series valve) at an inlet pressure of 20 & 45 psi:

In progress …

1737877110458.png

Brooks instrument

NRS series (8, 3 or 0.2 ℓ/min)


1737877156369.png
Brooks instrument 8513 (with digital turner handle)

  • Web: brooksinstrument.com
  • Part number (models): 8503/8504, 8513/8514
  • Orifice: n/a
  • Cv:
    • type 5: 0.017
    • type 4: 0.006
    • type 3: 0.0015
    • type 2: 0.0009
    • type 1: 0.0004
  • Price:
    • official: ?
    • ebay: ?
  • PDF: brooks---nrs.pdf
  • Thread: female NPT
  • Shut-off capability: yes
    • bubble tight shut-off
Graph missing …
Flow table missing …
  • Flow (at full open): ~0.2, ~0.4, ~0.7, ~2.8 and ~8 ℓ/min
To be continued ...
 
Chell

CMV series (0.02 ℓ/min)


1737877582217.png
Chell instruments CMV-VFM-1-P-22
  • Part number:
    • CMV(S)-?FM-1-* (ultra fine flow = 0-20 sccm)
    • CMV(S)-?FM-2-* (fine flow = 0-1000 sccm)
    • CMV(S)-?FM-3-* (medium flow = 0-3000 sccm)
  • Orifice: n/a
  • Cv: n/a
  • Flow (at 1 bar = ~15 psi):
    • 1st turn = 0.1 sccm (mℓ/min) = 3 bubbles per minute (bpm)
    • 5th turn = 1.0 sccm (mℓ/min) = 30 bpm
    • 7th turn = 2.0 sccm (mℓ/min) = 60 bpm
    • 20th turn (full open) = 20 sccm (mℓ/min) = 600 bpm = 10 bps
  • Price:
    • official: $670-700 (?)
    • ebay: ~$200
  • PDF:
  • Thread:
    • 1/16", 1/8", 1/4" or 6mm compression fittings
    • KF-10 vaccum fittings
    • welded VCR fittings
      • VCR is a special connector to which the hose/tubing can be attached (so you don't have to buy any additional connector)
  • Shut-off capability: yes
Graph of gas flow vs. valve handle turns (for Chell CMV series valve ultra fine flow) at an inlet pressure of 15 psi:

1737877717663.png

Flow (at 15 psi):

1737877741373.png
  • mℓ/min = sccm (cm3/min)
  • 1⁄20 = 1⁄20 of the 1st turn (the smallest possible fraction of one turn)
  • bps = bubbles per second (Ø 4 mm)
  • bpm = bubbles per minute
Aalborg

MFV Barstock valves (0.2 ℓ/min)


1737877784392.png
  • Web: aalborg.com
  • Part number:
    • VM1-BB-, VM1-SV- (Cv: 0.0005)
    • VM2-BB-, VM2-SV- (Cv: 0.001)
  • Orifice: 0.042" (1.1 mm)
  • Price:
    • official: $130
    • ebay: [not offered]
  • PDF: aalborg---mfv.pdf
  • Shut-off capability: yes
    • bubble tight shut-off
Graph missing …
Flow table missing …
  • Flow (at full open): ~0.2 and ~0.5 ℓ/min
    • according to the technical documentation: 6 and 12 mℓ/min (but that was calculated at different inlet pressure than other valves)
To be continued ...
 
Good (medium flow) needle valves

Hoke

2300 series [1° angle stem] (6 ℓ/min)
1737878086913.png

Graph of gas flow vs. valve handle turns (for Hoke 2300 series valve):

1737878025438.png

Flow (at full open): ~5.7 ℓ/min

Ideal valve

52/54-1 series (9 ℓ/min)


1737878117806.png
Ideal valve P52-1-11


Graph of gas flow vs. valve handle turns (for Ideal valve 52/54 series valve):

1737878217783.png

1737878233289.png

Flow (at full open): ~9 ℓ/min

To be continued ...
 
Last edited:
Swagelok (Nupro)

BM series [belows sealed] (9 ℓ/min)


1737878403233.png
Swagelok SS-4MG2
  • Web: swagelok.com
  • Part number: SS-4BMG, SS-4BMW (weld port)
  • Orifice: n/a
  • Cv: 0.019
  • Turns to full open: 6
  • Price:
    • official: ?
    • ebay: ?
  • PDF: swagelok---bm.pdf
Graph of gas flow vs. valve handle turns (for Swagelok BM series valve):

1737878452560.png

Flow (at full open): ~9 ℓ/min

Swagelok

M series (14 ℓ/min)


1737878482461.png
Swagelok SS-4MA-MH
  • Web: swagelok.com
  • Part number: SS(B)-?MG*, SS(B)-?MA*
  • Orifice: 0.056" (1.4 mm)
  • Cv:
    • 0.03
    • 0.026 (double pattern, dual handles)
  • Turns to full open: 9
  • Price:
    • official: ?
    • ebay: ?
  • PDF: swagelok---s-m-31.pdf
Graph of gas flow vs. valve handle turns (for Swagelok M series valve):

1737878552352.png

Flow (at full open): ~12.3 and ~14.2 ℓ/min

To be continued ...
 
[Hardly] satisfactory (higher flow) needle valves

Fabco

NV55/FC55 (19 ℓ/min)


1737878730620.png
Fabco NV-55
  • Web: fabco-air.com
  • Part number: NV-55, NV-55-18, FC-55
  • Orifice: n/a
  • Cv: ~0.04?
    • similar to Swagelok 31 series
  • Turns to full open: 9
  • Price:
    • official: $40-60
    • ebay: ?
  • PDF: fabco---needle-valves.pdf
Graph of gas flow vs. valve handle turns (for Fabco NV/FC series valve):

1737878800982.png

Flow (at full open): ~19 ℓ/min

Swagelok

31 series (19 ℓ/min)


1737878825413.png
Swagelok SS-31RS6MM
Graph of gas flow vs. valve handle turns (for Swagelok 31 series valve):

1737878876883.png

Flow (at full open): ~19 ℓ/min

To be continued ...
 
SMC penumatics

AS1000 series (20 ℓ/min)


1737879047147.png

Graph of gas flow vs. valve handle turns (for SMC AS series valve):

1737879103925.png

Flow (at full open): ~20 ℓ/min

Parker Hannifin

HR series (15 ℓ/min)


1737879135540.png
Parker Hannifin 2A-H5A-V-B-TC

  • Web: parker.com
  • Part number:
    • type 5: -H5?--- (Cv: 0.049)
    • type 4: -H4?--- (Cv: 0.032)
  • Orifice: n/a
  • Turns to full open: 15 (±1)
  • Price:
    • official: ?
    • ebay: ?
  • PDF: parker---n+hr-series.pdf
Graph of gas flow vs. valve handle turns (for Parker HR series valve):

1737879210975.png

Flow (at full open): ~15.2 and ~23.3 ℓ/min

NM series (26 ℓ/min)

Graph of gas flow vs. valve handle turns (for Parker NM series valve):

1737879272384.png

Flow (at full open): ~26 ℓ/min

Clippard

MNV-3/4 series (28 ℓ/min)


1737879316906.png
Clippard MNV-1K
  • Web: clippard.com
  • Part number: MNV-3*, MNV-4*
  • Orifice:
    • MNV-3: 0.07" (1.8 mm)
    • MNV-4: 0.067" (1.7 mm)
  • Cv: n/a
    • at 4 turn SCFM is 1, Cv is under 0.03 before 4 turns; after 4 turns, Cv spike
  • Turns to full open: ?
  • Price:
    • official: ?
    • ebay: ?
  • PDF: clippard---mechanical-valves.pdf
Graph missing …
Flow table missing …

  • Flow (at full open): 28 ℓ/min
To be continued ...
 
Unsatisfactory (high flow) needle valves

Pneumadyne

700 series (43 ℓ/min)


1737879570600.png
Pneumadyne C070501

Graph of gas flow vs. valve handle turns (for Pneumadyne 700 series valve):

1737879624928.png

Flow (at full open): ~43 ℓ/min

Swagelok

20 series (43 ℓ/min)


1737879662852.png
Swagelok SS-20VS4
  • Part number: SS-20R*, B-20R*
  • Orifice: 0.08" (2.0 mm)
  • Cv: 0.09
  • Turns to full open: 2.5
  • Price:
    • official: ---
    • ebay: $60-100
Flow (at full open): ~43 ℓ/min

O series (43 ℓ/min)

1737879713484.png
Swagelok SS-ORS2

  • Part number: SS-OR*, B-OR*
  • Orifice: 0.08" (2.0 mm)
  • Cv: 0.09
  • Turns to full open: 8
  • Price:
    • official: ?
    • ebay: ?
Flow (at full open): ~43 ℓ/min

Camozzi

RFU/RFO series (50 ℓ/min)


1737879752298.png
Camozzi RFO 352-M5


  • Web: camozzi.com
  • Part number: RFU 452-M5, RFO 352-M5
  • Orifice: 0.06" (1.5 mm)
  • Cv: 0.116 (?)
  • Turns to full open: 17
  • Price:
    • official: $15-20 (?)
    • ebay: ?
  • PDF: camozzi---rfu-rfo.pdf
  • Thread: M5, 1/8" G, 1/4" G
  • Shut-off capability: no
    • The handle stop is set at ~3 ℓ/min (!) with 87 psi (6 bar) inlet pressure. This means that this valve cannot be set to a flow rate lower than 3 ℓ/min (= 1,500 bps). Adjusting stop to lower flow setting can damage valve and stem tip!
Graph of gas flow vs. valve handle turns (for Camozzi RFU/RFO series valve) at an inlet pressure of 87 psi (6 bar):

1737879903878.png

Flow (at 87 psi):

1737879925008.png
  • mℓ/min = sccm (cm3/min)
  • 1⁄20 = 1⁄20 of the 1st turn (the smallest possible fraction of one turn)
  • bps = bubbles per second (Ø 4 mm)
  • bpm = bubbles per minute
Most hobby-grade valves
  • most unbranded/no-name valves used as part of cheap CO2 sets
[End of article]
 
Last edited:
@Bettatail was very helpful to me too when I sought out a nice Swagelok NV. He even custom made a fitting for me so that it would fit my setup after I advised him of what I had.

Good article
 
NOTICE: Given my severely limited ability to communicate with others via social media in today's world in a way that is mutually agreeable and does not lead to conflict (I've tried many times and always failed), I prefer not to participate in any online discussion. If you have any questions or would like to discuss other matters with me, please use the mail channel: golias.net/akvaristika/contact.php.
 
Thanks Marcel, for putting the info out here.

This is bettatail from TPT, I am here because I was informed by a couple planted tank hobbyists that my metering valves selection info is in UK scape/planted tank forum.
The linked list of metering valves selection was post on 2011(last edited in 2014). I made a new selection thread in 2020, here.

a bubble double in diameter, the volume of the bubble is 8 times of the original. There are many factors affect the bubble size, so steady bubbles only show you a relatively slow or fast steady co2 flow rate, but not the actual flow rate or volume---- I did some research on the high precision glass tube flow meters and introduced them in my 2020 metering valve selection thread, and started the discussion about the floating ball flow meters in a couple more threads. you can check THIS here.
Mass flowmeters are the prevailing high precision flow control devices, have long replaced the glass tube flow meters. I don't talk about the mass flowmeters because they are out of reach for any average planted tank hobbyist.

The metering valves selection thread was a guide for planted tank hobbyists who want to DIY their own high end co2 system using industrial/lab grade parts, in the United States. Of course, there are three major components for a co2 system-- the regulator, solenoid and needle/metering valve. The discussions about the different low flow, low power consumption solenoids and high purity/heavy duty double stage regulators had never generated a list of brands/model number or consolidated into a single thread or post as metering valves, they are scattered here or there on TPT.

The topic of DIY high end co2 system was popular back then, because it was during the years after 2008 market crash and production outsourcing to China, there were many liquidated/used industrial grade parts on US ebay. In those years, it was a feast for planted tank hobbyists in the US, many spent only a fraction of the original cost to get these high value, high quality parts and made their own high end co2 systems for their planted tanks.
Now in 2025, any hobbyist from the US who wants to make his own set really need some luck to find all three major components in good price, and for people outside of US, chance of finding cheap and high quality parts is weaker, even after US ebay open to most of the countries now a day.
 
In my spare time (of which there isn't much right now) I'm working on a practical guide to choosing the right needle valve for a particular setup.
Here is a small sample (which includes only a few selected needle valves so far):

1743936123039.png

I would then like to convert this into some interactive wizard on my website where you would just enter selected input values (i.e. desired bps + gauge pressure used), and it would automatically show you the appropriate (recommended) valves.
 
Last edited:
I went to the trouble of recalculating the maximum CO2 flow for each needle valve. The technical documentation usually lists the maximum gas flow when the valve is fully open, and these figures are measured at different pressures by different manufacturers, so they don't compare very well with each other. So I normalised the figures (where I could) to give the gas flow not at full valve opening but at one turn of the valve knob, and I also converted it from the meaningless litres per minute (or CFM) to bubbles per second (= bps). [I used 4 mm diameter bubbles in my calculations.] So the figures below show how many bps the CO2 flow through the valve will increase by if you open it one turn (= 360°). Since the flow rate is not linear, but usually exponential, it will not increase by a factor of two, but by a factor of several (which is good to keep in mind) when you open it by two turns, etc. When assessing the (un)suitability of individual valves for aquarium purposes, it is also good to remember that the approximate flow rate of CO2 in our aquariums [to achieve a target concentration of around 30 ppm] is about 1 bps per 60L. So if you have a 60L aquarium and you want to have 30 ppm CO2 in there, you only need to add about 1 bps (assuming it dissolves perfectly). In a 180L aquarium, you will need to add about 3 bps of CO2. Of course, these numbers are just a guideline, but they will help you in judging which valve is suitable for our purposes and which is less so. As you will see, the vast majority of valves are unsuitable because they are designed for completely different (several times higher) flow rates.

Example:
The Hoke Micromite 1600 needle valve will give you a 2.4 bps increase in CO2 flow with one turn of the handle (t), which is excellent. This valve is therefore ideal for CO2 control in the aquarium. In contrast, the Fabco NV-55 needle valve will increase your CO2 flow by 951 bps at one turn of the handle, which is a disastrous value. This valve is therefore unsuitable for CO2 control in the aquarium. If you wanted to set a value of e.g. 2 bps with this valve, you would have to turn the valve handle so slightly that there is no question of precision or fine control. CO2 control with this valve is basically trial and error. Although in practice the resulting regulation is little better than "on paper", the data below will hopefully give you a rough idea of what valves are really suitable for our purposes.

needle-valves---update.png
Here's a link to the same picture in its original resolution: golias.net/akvaristika/img/valves/needle-valves---update.png

PS: Unfortunately, I do not have enough data to calculate the bps for the Brooks NRS and Clippard MVN valves. However, I will try to at least estimate them later.
 
Last edited:
Just get a Mass flow meter or a glass tube flow meter to measure the ACTUAL flow volume on each ACTUAL needle/metering valve, at low flow end which around CV=0.00025.
 
Thank you @Marcel G for this great thread.

Perhaps one take away is that most commercial needle valves are generally designed for much higher flows than we need in our tanks, and that we are operating them on the (lower) limits of what they can handle reliably?

Some may disagree, but I believe that CO2 stability is what mostly counts and I am not really interested in small adjustments. Who knows if 30 ppm CO2 is better than 20% more or 20% less than that? Also, I have never seen anyone describing what "dialling in CO2" really means and how we do it.

Do we really need to adjust CO2, "dial in" as we like to say in our hobby, or do we want to target a reasonable flow rate and then keep this rock solid constant from then on? In other words, if we had a needle valve that once it has been set at a decent flow rate could be locked and stable forever?

What would be the pro's and con's of some cheap gas nozzles (just a hole in a piece of metal), with calibrated hole sizes to meet our CO2 flow target? Example in case may be diesel engine fuel injectors, which can have really small holes (80 micrometers and smaller) and can be mass produced and purchased at reasonable price.

Ideally we could calculate which injector we would need for our size tank, and CO2 ppm target, and apply small adjustments if needed by setting the pressure from the dual stage regulator.

If it works we could get stability for a good price, and would not have to worry anymore about (semi) professional regulators that we may use outside the zone they were designed for.

I hope that someone has the tools and skills to experiment, using commercially available fuel injectors, and see if it works in practice.
 
Perhaps one take away is that most commercial needle valves are generally designed for much higher flows than we need in our tanks, and that we are operating them on the (lower) limits of what they can handle reliably?
Yes, that's quite obvious from the above data set.
Some may disagree, but I believe that CO2 stability is what mostly counts and I am not really interested in small adjustments. Who knows if 30 ppm CO2 is better than 20% more or 20% less than that? Also, I have never seen anyone describing what "dialling in CO2" really means and how we do it. Do we really need to adjust CO2, "dial in" as we like to say in our hobby, or do we want to target a reasonable flow rate and then keep this rock solid constant from then on? In other words, if we had a needle valve that once it has been set at a decent flow rate could be locked and stable forever?
100% agreement. The term "dialing in" or "tuning CO2" is, in my opinion, just an attempt to make an art out of what should be a simple task. If you use a suitable needle valve (such as a Hoke Micromite 1600) to set the desired CO2 flow, setting the correct flow is a piece of cake. The only problem is to find out if the set flow rate matches the desired target CO2 concentration (e.g. 30 ppm), which is what the "drop checker" is used for, but the color change of its solution normally has a delay of several hours compared to reality, and therefore the specific flow rate setting (= bps) on the needle valve then needs to be "fine-tuned" (= adjusted) so that the specific flow rate matches the desired CO2 concentration (i.e. so that the result is a green drop checker = 30-35 ppm CO2). Of course, if you use an unsuitable valve to adjust the flow, then it really becomes an art, because hitting, for example, 2 bps on a valve that gives you maybe 10 times that even with the slightest turn may not be easy (but it is certainly possible, as confirmed by common practice => unsuitable valves are certainly used by the vast majority of aquarists, as they are used in the vast majority of commercial CO2 sets).
What would be the pro's and con's of some cheap gas nozzles (just a hole in a piece of metal), with calibrated hole sizes to meet our CO2 flow target? Example in case may be diesel engine fuel injectors, which can have really small holes (80 micrometers and smaller) and can be mass produced and purchased at reasonable price. Ideally we could calculate which injector we would need for our size tank, and CO2 ppm target, and apply small adjustments if needed by setting the pressure from the dual stage regulator. If it works we could get stability for a good price, and would not have to worry anymore about (semi) professional regulators that we may use outside the zone they were designed for. I hope that someone has the tools and skills to experiment, using commercially available fuel injectors, and see if it works in practice.
This seems to me to be a very good idea (an innovation worth testing!). The only problem I can think of now in connection with this might be the different dissolution efficiency of CO2 in different aquariums. Because each injector will be sized for a specific flow rate (= to reach a specific target CO2 concentration). It's just that different aquarists use different methods of dissolving CO2 => some 100% efficient, others much less efficient. But if the injectors were cheap enough and could be put into some universal "cartridge" for example, then every aquarist could get some "injector kit" and simply change them on the fly. Then it wouldn't matter what method of CO2 dissolution one uses. Simply put e.g. injector #1 (= 80 μm diameter) in the cartridge and in a few hours you would see what colour the drop checker in your aquarium would be. If the resulting CO2 concentration was too low, you would simply replace the injector #1 with the injector #2 (= 100 μm), etc. Replacing the preset injectors would then essentially replace turning the knob on the needle valve. The main issue here is the cost of such a kit. A good quality needle valve can be found on ebay in the $50-100 range (if one is lucky). Nobody reasonable will buy a new one (which costs $300+). So this would have to cost under $50 or so.

PS: Basically it would be the same principle as your "CO2 reactor with overflow". There, you also have to make a certain set of tubes (or vessels) because each tube dissolves a certain amount of CO2 in the aquarium water. And if you have a bigger aquarium, you have to use a bigger tube.
 
Just get a Mass flow meter or a glass tube flow meter to measure the ACTUAL flow volume on each ACTUAL needle/metering valve, at low flow end which around CV=0.00025.
Flow meters are expensive and if the above needle valves have such a large variance in flow rates, I would have to buy several flow meters (each for a certain range of flow). Also, flow meters are hard to find in Europe. Many ebay sellers in the US or Canada do not send them to Europe. In addition, it has been my experience that many valves that are the same on paper give different results in practice. For example, some valves can completely stop the flow, whereas others, even when fully closed, will still let a small amount of gas through (which is normal according to the manufacturers, as most are not intended to completely shut off the gas supply, but only to regulate the flow within a certain range). No two valves of the same make/model will give you the same results in practice. Additionally, many valves purchased second hand may have the needle damaged in some way (even minimally) or be slightly clogged, which can also affect the results. Therefore, I find it safer to go by data measured by professionals at a testing institute. Just use common sense on this then and bear in mind that in practice it may vary slightly (but not by an order of magnitude) for each particular valve. My aim here is not to provide any absolutely precise figures, but rather to find out how low or high flow rates can actually be expected from individual valves. My contribution is that I have tried to normalize all the results so that they are comparable to each other (which, as far as I know, no one has done yet). I'm pretty busy with work right now, so I unfortunately don't have time to detail my methodology here, but I plan to write an article about it in the summer where I'll try to explain it in more detail. For most of the valves I have already been able to create my own graphs of flow rate vs. pressure and valve handle turns, so everyone will then be able to calculate exactly what flow rate they can expect at different gauge pressures (psig) for each valve. [My friends already have access to most of it on my site.] I personally find the use of needle valves in aquaristics to be very problematic and so prefer to look for other (more appropriate) methods of CO2 application, but since I have delved into it in an effort to understand it, and hopefully understand it, and because I am now able to do the math and convert it to our "bubbles per second" (or other metrics), I plan to share it then so that it doesn't just stay with me and others can possibly benefit from it (if they are interested). But of course, everyone is free to do with it as they see fit. I mean, he can even criticize it as a complete wastage.

PS: Most aquarists will certainly argue that they have been using Fabco valves for years and have no problem setting even extremely low flow rates (e.g. 1-3 bps) without understanding the point of my post and their inherent unsuitability. I sometimes liken it to a mountain road. If you want to take a car up a mountain, then the appropriate car for that purpose is an off-road car, not a city car (i.e. Citigo ForTwo). However, you will certainly find plenty of people on the internet trying to prove to you how they managed to negotiate some difficult terrain in an ordinary Citigo. The same thing happens in aquaristics, of course. Those who use Fabco valves will try to convince you that nothing better is needed and that they can set the required flow rate with a little skill. These people don't understand the difference between (un)suitable and (un)possible.
 
Last edited:
Back
Top