Arc minutes are a rate, not a grade
Sensitivity on a bubble vial is defined as the angle you must tilt it through to move the bubble a set distance, and that distance is conventionally 2 mm. So 30' does not mean the vial is "accurate to 30 minutes". It means the bubble travels 2 mm for every 30 minutes of arc you tilt it through.
That is a rate. To use it you have to convert it into something you can see on the job: millimetres of height per metre of base. Thirty minutes of arc is 0.5 of a degree, and over a metre that is 8.73 mm.
This article converts all nineteen item numbers in this series into that form, then does the thing the parameter table does not: it works back to the number the factory has to hold, which is the radius of the bore.
The series converted into working units
The published table gives item number, diameter, length and sensitivity. The last three columns are derived: sensitivity in millimetres per metre, the same figure in inches per foot, and the bore radius that produces it.
| Item No. | Dia. mm | L mm | Sensitivity | mm/m | in/ft | Bore radius mm |
|---|---|---|---|---|---|---|
| RB-CY0512 | 5 | 12 | 30' ±6' | 8.73 | 0.105 | 229 |
| RB-CY0513 | 5 | 13 | 45' ±6' | 13.09 | 0.157 | 153 |
| RB-CY0614 | 6 | 14 | 30' ±6' | 8.73 | 0.105 | 229 |
| RB-CY0615 | 6 | 15 | 45' ±6' | 13.09 | 0.157 | 153 |
| RB-CY0620 | 6 | 20 | 45' ±6' | 13.09 | 0.157 | 153 |
| RB-CY0629 | 6 | 29 | 45' ±6' | 13.09 | 0.157 | 153 |
| RB-CY6518 | 6.5 | 18 | 20 to 25' | 5.82 to 7.27 | 0.070 to 0.087 | 306 to 382 |
| RB-CY0718A-30 | 7 | 18 | 30' ±6' | 8.73 | 0.105 | 229 |
| RB-CY0718B-60 | 7 | 18 | 60' ±6' | 17.46 | 0.210 | 115 |
| RB-CY77254 | 7.7 | 25.4 | 45' ±6' | 13.09 | 0.157 | 153 |
| RB-CY7726 | 7.7 | 26 | 45' ±6' | 13.09 | 0.157 | 153 |
| RB-CY0727 | 7 | 26.4 | 30' ±6' | 8.73 | 0.105 | 229 |
| RB-CY0730 | 7 | 30 | 60' ±6' | 17.46 | 0.210 | 115 |
| RB-CY0820 | 8 | 20 | 30' ±6' | 8.73 | 0.105 | 229 |
| RB-CY0822A-30 | 8 | 22 | 30' ±6' | 8.73 | 0.105 | 229 |
| RB-CY0822B-80 | 8 | 22 | 80' ±10' | 23.28 | 0.279 | 86 |
| RB-CY0822C-120 | 8 | 22 | 120' ±10' | 34.92 | 0.419 | 57 |
| RB-CY0823A-30 | 8 | 22 | 30' ±6' | 8.73 | 0.105 | 229 |
| RB-CY0823B-60 | 8 | 22 | 60' ±6' | 17.46 | 0.210 | 115 |
The radius comes from the definition itself. If the bubble moves 2 mm for an angle of theta radians, then the bore it rides in has radius R = 2 / theta. Thirty minutes gives 229 mm; 120 minutes gives 57 mm.
That single column is the most useful thing on the page, because radius is what the mould or the grind has to produce. Everything else in the table is a restatement of it.
Two vials from one batch can differ by 50%
The tolerance is published, which is better than most catalogues manage, and it is worth reading carefully before quoting a number to a customer.
| Nominal | Band in mm/m | Relative error | Worst two vials in a batch |
|---|---|---|---|
| 30' ±6' | 6.98 to 10.47 | 20% | differ by 1.50 times |
| 45' ±6' | 11.34 to 14.85 | 13% | differ by 1.31 times |
| 60' ±6' | 15.71 to 19.20 | 10% | differ by 1.22 times |
| 80' ±10' | 20.35 to 26.18 | 12% | differ by 1.29 times |
| 120' ±10' | 32.00 to 37.84 | 8% | differ by 1.18 times |
Note what the tolerance does as the nominal gets coarser. Every ±6' row carries the same absolute band, 1.75 mm/m, so the relative error falls from 20% at 30' to 10% at 60'. The coarse vials are relatively more honest than the fine ones.
The practical consequence: a 30' vial at the bottom of its band and another at the top differ by a factor of 1.5. If you are building a machine that is levelled against its own vial, that is fine, because the vial is a reference to itself. If you are shipping vials as spares and expecting one to replace another, it is not, and you should say so in the purchase order.
The number the factory actually has to hold
A bore of radius R running along a body of length L is not straight, it is an arc. The rise from one end to the middle is the sagitta, h = L squared / (8R).
| Item | L mm | Sensitivity | Bore radius mm | Bore sagitta mm |
|---|---|---|---|---|
| RB-CY0512 | 12 | 30' | 229 | 0.079 |
| RB-CY0614 | 14 | 30' | 229 | 0.107 |
| RB-CY6518 | 18 | 22.5' | 306 | 0.133 |
| RB-CY0513 | 13 | 45' | 153 | 0.138 |
| RB-CY0615 | 15 | 45' | 153 | 0.184 |
| RB-CY0820 | 20 | 30' | 229 | 0.218 |
| RB-CY0629 | 29 | 45' | 153 | 0.688 |
| RB-CY0730 | 30 | 60' | 115 | 0.982 |
| RB-CY0822C-120 | 22 | 120' | 57 | 1.056 |
Look at the top of that table. The 5 mm by 12 mm vial at 30' has a bore that rises 0.079 mm over its entire length. The 6 mm by 14 mm at the same sensitivity has 0.107 mm.
Normal tolerance on an injection moulded acrylic part is on the order of 0.1 mm. So for the shortest and finest vials in this series, the curvature that defines the sensitivity is at or below the tolerance of the process that makes it.
That is not a criticism of the product; it is the reason these vials exist in a range of lengths. A sensitivity cannot be formed reliably if the whole arc is smaller than the mould variation. Either the bore is ground, or the body is long enough to put the arc above the noise, or the vial is sorted after moulding and the sensitivity is a measured outcome rather than a moulded feature.
Which of those three it is decides what you are buying. It is the first question in the list at the end.
Coarse is the entire point
It is easy to read "30 minutes" and assume it is a precision instrument. It is not.
A machinist's level at 0.02 mm/m is the reference grade: that is one division for 0.02 mm of rise over a metre, which equals 0.069 arc minutes, or 4.1 arc seconds.
Thirty minutes is 436 times coarser than that. Forty-five minutes is 654 times, and 120 minutes is 1,745 times.
These vials are not for scraping a machine bed flat. They are for telling an operator that a frame, a tripod, a camera shoe, a 3D printer bed or a cabinet is close enough to level. Judged by that job, 8.7 mm per metre is a sensible figure: it is visible, it settles quickly, and it does not chase vibration.
What one division looks like on a real base
To move the bubble 2 mm, you tilt the vial by its stated angle. On a base of known length that tilt is a height difference you can measure with a shim.
| Base length | 30' | 45' | 60' | 120' |
|---|---|---|---|---|
| 100 mm | 0.87 mm | 1.31 mm | 1.75 mm | 3.49 mm |
| 200 mm | 1.75 mm | 2.62 mm | 3.49 mm | 6.98 mm |
| 300 mm | 2.62 mm | 3.93 mm | 5.24 mm | 10.48 mm |
| 500 mm | 4.36 mm | 6.55 mm | 8.73 mm | 17.46 mm |
| 1,000 mm | 8.73 mm | 13.09 mm | 17.46 mm | 34.92 mm |
So on a 200 mm base, one division of a 30' vial is 1.75 mm of shim under one end. On a 1,000 mm base it is 8.73 mm.
This is also the incoming inspection. Put the vial in a known flat seat, raise one end of the base by the shim from the table, and the bubble should move one division. A feeler stack gets you to 0.05 mm, which on a 200 mm base is 3% of the angle, well inside the ±20% tolerance you are checking against.
Do the same test on two vials from the same carton and you can see for yourself whether the batch is consistent, without a certificate.
Sensitivity is not range
This is the part the tables never cover, and it catches people out.
The bubble can only travel so far before it pins against the end of the barrel. That distance is set by the barrel length minus the bubble length, halved. Divide by 2 mm and you have the number of divisions the vial can actually show.
The table below assumes 1.5 mm of end cap at each end and a 2.5 mm bubble. Both are assumptions, because neither number is published; they are here to show the shape of the problem, not to replace a measurement.
| Item | Body L | Sensitivity | Barrel, assumed | Bubble travel | Tilt it can show |
|---|---|---|---|---|---|
| RB-CY0512 | 12 mm | 30' | 9 mm | ±1.6 divisions | ±49' = 14.2 mm/m |
| RB-CY0620 | 20 mm | 45' | 17 mm | ±3.6 divisions | ±163' = 47.6 mm/m |
| RB-CY0822A-30 | 22 mm | 30' | 19 mm | ±4.1 divisions | ±124' = 36.1 mm/m |
| RB-CY0822C-120 | 22 mm | 120' | 19 mm | ±4.1 divisions | ±496' = 145 mm/m |
The 5 mm by 12 mm vial can show about 1.6 divisions. At 30' each, that is roughly ±49 minutes, or ±14 mm/m. Beyond that the bubble is against the end and the vial says only "far off".
The 8 mm by 22 mm vial can show about 4.1 divisions. At 30' that is ±124 minutes; at 120' it is ±496 minutes, which is over eight degrees.
So a short fine vial is a go or no-go gauge, and a long coarse vial is a measuring instrument. Choosing the finest sensitivity in the shortest body gives you the least usable device in the catalogue, which is the opposite of what the number suggests.
The two printed lines are the whole scale
The photographs show each vial with two black lines printed around the barrel, with the bubble sitting between them. There is no graduated scale, no numbers, and no centre mark.
That is normal for this class of vial and it is not a defect. But it defines what you can read. You can see: bubble inside the lines, bubble touching a line, bubble past a line. You cannot read a value off it.
It also means the line spacing is the de facto division, and it is not published. If the lines are 2 mm apart and the definition of sensitivity is 2 mm of travel, the two agree. If they are 3 mm apart, then the centre-to-line displacement is 1.5 mm, or three quarters of a division, and the vial reads differently to the specification.
For a vial sold by its sensitivity, the line spacing should be stated on the drawing. It is not, and that is worth one line in the enquiry.
Part numbers that break their own rule
Most of the series follows a readable rule: the code is the diameter with the decimal point removed, then the length with the decimal point removed.
| Item No. | Reads as | Table says | Agrees |
|---|---|---|---|
| RB-CY0512 | 5.0 mm dia, 12 mm long | 5 / 12 | yes |
| RB-CY6518 | 6.5 mm dia, 18 mm long | 6.5 / 18 | yes |
| RB-CY77254 | 7.7 mm dia, 25.4 mm long | 7.7 / 25.4 | yes |
| RB-CY0820 | 8.0 mm dia, 20 mm long | 8 / 20 | yes |
| RB-CY0823A-30 | 8.0 mm dia, 23 mm long | 8 / 22 | no |
| RB-CY0727 | 7.0 mm dia, 27 mm long | 7 / 26.4 | no |
Four of the six agree. Then two do not:
RB-CY0823A-30 and RB-CY0823B-60 both list a length of 22 mm, but the code carries 23. Under the rule they would be RB-CY0822A and RB-CY0822B, and one of those numbers already exists.
RB-CY0727 lists 26.4 mm while the code carries 27. Under the rule it would be RB-CY07264. 26.4 mm is also an unusual figure for a moulded body, so it may be the table that is wrong rather than the code.
Either way, ordering by code and ordering by dimension give two different parts for those three item numbers. Put the dimension on the purchase order as well as the code.
Two part numbers, one specification
The sharpest problem in the table:
| Field | RB-CY0822A-30 | RB-CY0823A-30 |
|---|---|---|
| Diameter | 8 mm | 8 mm |
| Length | 22 mm | 22 mm |
| Sensitivity | 30' ±6' | 30' ±6' |
| mm/m | 8.73 | 8.73 |
| Bore radius | 229 mm | 229 mm |
Identical in every published field, with different codes. Something distinguishes them, because a manufacturer does not carry two part numbers for one part. The candidates are bubble length, line spacing, body colour or material, or one of them simply being a discontinued code still printed in the table.
Until that is answered, treat the two as interchangeable only if you can accept either. If your assembly has a fixture cut for one, ask which one is current.
There is a second pattern worth noting: 19 item numbers describe 18 distinct combinations of diameter, length and sensitivity. The only duplication is the pair above.
One imperial vial in a metric catalogue
RB-CY77254 is 7.7 mm by 25.4 mm. That is 0.303 inch by 1.000 inch, exactly.
A round 25.4 mm is not a coincidence in a series where every other length is a whole millimetre. This is an inch-series vial that has been relabelled in metric for this catalogue, which is common and harmless, but it does mean the tolerance on that one is probably held to an imperial drawing. If you are matching an existing US-made part, that is the one to look at first.
What the photographs do not show
Two of the eight images show the bare vials, in blue and in yellow-green, with the two printed lines and rounded end caps. Those are useful.
The rest are a catalogue collage of about thirty finished level products in housings, a set of packing crates, a shipping and payment panel, a factory gate, an FAQ sheet and a business card.
None of them is a dimension drawing. There is no view of a vial with the diameter, the length, the line spacing or the end cap called out, and no close-up of a cap joint. For a component whose whole specification is geometry, that is the gap that matters, and it is the same gap in almost every vial catalogue. Ask for the drawing rather than expect to find one.
What is genuinely strong in this series is the table itself: nineteen rows, sensitivity given to the arc minute with a tolerance beside it. Most vial catalogues stop at diameter and length and leave sensitivity as a single word. That table is worth more than the photographs, and it is enough to specify from.
Which of the three groups to start from
The series sorts itself by diameter, and the sorting is not arbitrary.
5 and 6 mm only ever come at 30' or 45'. At those sizes the bubble is a large fraction of the bore, so the vial reads as a clear in-or-out signal, and the fine sensitivities keep the arc manufacturable relative to the small body.
7 mm is the middle ground: four items, 18 to 30 mm long, carrying 30' or 60'. If you want one vial that both reads and measures, this is where to look.
8 mm carries everything above 60'. The 80' and 120' vials exist only at 8 mm by 22 mm, because a coarse sensitivity needs a large bore and enough length to make the arc, and because a fast-settling bubble wants volume.
So: fine and short for a compact assembly that only needs to confirm level; coarse and long where an operator has to bring something into level from visibly off.
Fourteen questions for the quotation
The table is strong enough to specify from. These are the gaps that remain, and the first one is structural.
| Question | Why it decides the order |
|---|---|
| Is Dia. the outside diameter or the bore | The whole table rests on this, and it is never stated |
| What is the spacing of the two printed lines | The 2 mm division is the definition of sensitivity; the lines are the only scale on the part |
| What is the bubble length | It sets how much travel is left before the bubble pins at the end |
| What is the usable barrel length between the end caps | Together with bubble length it fixes the measuring range |
| What differs between RB-CY0822A-30 and RB-CY0823A-30 | Every published field is identical |
| Is RB-CY0823A-30 really 22 mm or 23 mm long | The code and the table disagree |
| Is RB-CY0727 26.4 mm or 27 mm long | Same problem, and 26.4 is an odd number for a moulded body |
| Is the 20 to 25' on RB-CY6518 a range or a tolerance | Every other row states a nominal with a tolerance |
| Body material, acrylic or glass | Not in the table; acrylic and glass differ in scratch resistance and in how they age |
| Fluid type, dye and low-temperature behaviour | Viscosity sets how fast the bubble settles; no temperature range is published |
| At what temperature the sensitivity is measured | A sensitivity figure without a reference temperature is not reproducible |
| Is the bore moulded or ground | The finest short vials need less than 0.1 mm of sagitta, which is inside normal moulding tolerance |
| Batch sensitivity certificate, measured | The only way to know the vial you get is the one you specified |
| End cap construction | Solvent bond, ultrasonic weld and a pressed plug fail differently |
Where this fits with the other vial pages
The cylinder water vial selection guide covers material, colour and certification, which this table does not. The circular bubble vial guide covers the bullseye type, where sensitivity works differently because the bubble moves in two axes at once. This page is only about the arc-minute figure and what it costs you in geometry.
Frequently asked questions
What does 30 minutes mean on a bubble vial?
It is the angle of tilt needed to move the bubble 2 mm, which is the conventional division. Thirty minutes of arc is half a degree, so over a one metre base it corresponds to 8.73 mm of height difference. It is a rate of response, not an accuracy figure.
Is a 30 minute vial more accurate than a 120 minute one?
More sensitive, yes: it responds to a smaller tilt. But its tolerance is 20% of nominal against 8% for the 120 minute vial, so two 30 minute vials can differ by 1.5 times while two 120 minute vials differ by 1.18 times. Sensitivity and consistency are different things.
Why does the bore radius matter to a buyer?
Because it is the only number the factory actually has to produce. Radius equals 2 mm divided by the angle in radians, so 30 minutes means a 229 mm radius and 120 minutes means 57 mm. Everything else in the table is derived from it.
Why can a short fine vial not measure how far off level something is?
Because the bubble runs out of travel. A 5 by 12 mm vial has roughly 9 mm of barrel, so the bubble can move about 1.6 divisions before it pins. At 30 minutes each that is about 49 minutes, or 14 mm per metre. Past that the vial only says far off.
Can I check the sensitivity myself on receipt?
Yes, with a flat seat and a shim. On a 200 mm base, raise one end by 1.75 mm and a 30 minute bubble should move exactly one division. A feeler stack at 0.05 mm gives you 3% resolution on that test, which is well inside the 20% tolerance.
Why are two item numbers identical?
RB-CY0822A-30 and RB-CY0823A-30 share diameter, length, sensitivity and therefore every derived figure. Something unpublished distinguishes them, possibly bubble length or line spacing. Ask which is current before cutting a fixture.
Do these vials have a graduated scale?
No. Each has two printed lines with the bubble between them, which gives you in, touching or past. There is no numbered scale, so the line spacing is effectively the division and it should be stated on the drawing. It is not.
Key specifications
Frequently asked questions
- What does 30 minutes mean on a bubble vial?
- It is the angle of tilt needed to move the bubble 2 mm, which is the conventional division. Thirty minutes of arc is half a degree, so over a one metre base it corresponds to 8.73 mm of height difference. It is a rate of response, not an accuracy figure.
- Is a 30 minute vial more accurate than a 120 minute one?
- More sensitive, yes: it responds to a smaller tilt. But its tolerance is 20% of nominal against 8% for the 120 minute vial, so two 30 minute vials can differ by 1.5 times while two 120 minute vials differ by 1.18 times. Sensitivity and consistency are different things.
- Why does the bore radius matter to a buyer?
- Because it is the only number the factory actually has to produce. Radius equals 2 mm divided by the angle in radians, so 30 minutes means a 229 mm radius and 120 minutes means 57 mm. Everything else in the table is derived from it.
- Why can a short fine vial not measure how far off level something is?
- Because the bubble runs out of travel. A 5 by 12 mm vial has roughly 9 mm of barrel, so the bubble can move about 1.6 divisions before it pins. At 30 minutes each that is about 49 minutes, or 14 mm per metre. Past that the vial only says far off.
- Can I check the sensitivity myself on receipt?
- Yes, with a flat seat and a shim. On a 200 mm base, raise one end by 1.75 mm and a 30 minute bubble should move exactly one division. A feeler stack at 0.05 mm gives you 3% resolution on that test, which is well inside the 20% tolerance.
- Why are two item numbers identical?
- RB-CY0822A-30 and RB-CY0823A-30 share diameter, length, sensitivity and therefore every derived figure. Something unpublished distinguishes them, possibly bubble length or line spacing. Ask which is current before cutting a fixture.
- Do these vials have a graduated scale?
- No. Each has two printed lines with the bubble between them, which gives you in, touching or past. There is no numbered scale, so the line spacing is effectively the division and it should be stated on the drawing. It is not.