What a cased bullseye is, and why it replaces two vials
This is a turned aluminium flange with a circular vial set into the centre and three holes drilled round the pitch circle for mounting. The body runs from 20 mm to 60 mm across, the window from 14 mm to 43 mm, and the whole thing stands 10 to 16 mm tall.
The vial is not a barrel, it is a spherical cap. That single difference is why one of these does the job of two tubular vials: the bubble is a disc, and it moves radially, so a single instrument reads tilt in both axes at once. Two crossed tubes can measure the same two axes, but only after each has been aligned to the same reference, and that alignment is an error you add yourself.
What follows is about the part as a thing to buy: whether its numbers hold together, how to check them without a certificate, and what the geometry costs you once it is on a machine.
Eighteen items, in millimetres
The series runs from a 20 mm body to a 60 mm body in eighteen sizes. Every figure below is in millimetres: D1 is the outside diameter of the flange, D2 the window, H1 and H2 the two heights, PCD the pitch circle of the mounting holes, and the last column the sensitivity.
| Item | D1 body | D2 window | H1 | H2 | PCD | Hole | Sensitivity |
|---|---|---|---|---|---|---|---|
| HSSRX-M2010 | 20 | 14 | 10 | 2 | 17 | 2.0 | 10'/15'/30' |
| HSSRX-M2310 | 22.5 | 14 | 10 | 2 | 17.8 | 2.5 | 10'/15'/30' |
| HSSRX-M2512 | 25.5 | 16.5 | 11.5 | 2 | 21 | 2.5 | 30' |
| HSSRX-M2812 | 28 | 16 | 16 | 12 | 22 | 2.8 | 10'/15'/30' |
| HSSRX-M3011 | 30 | 20 | 11 | 3.5 | 25 | 3.2 | 10'/15'/30' |
| HSSRX-M3013 | 29.2 | 19.2 | 13 | 3 | 24 | 3.2 | 10'/15'/30' |
| HSSRX-M3212 | 32 | 20 | 12 | 4 | 27 | 3.2 | 10'/15'/30' |
| HSSRX-M3412 | 34 | 20 | 12 | 4 | 27 | 3.5 | 10'/15'/30' |
| HSSRX-M3812 | 38 | 28 | 12 | 3 | 37.5 | 3.5 | 10'/15'/30' |
| HSSRX-M3912 | 39 | 19 | 12 | 4 | 30 | 3.4 | 10'/15'/30' |
| HSSRX-M4013 | 40 | 30 | 13 | 4 | 34.5 | 3.0 | 30' |
| HSSRX-M4014 | 40 | 22 | 14 | 4 | 34.5 | 4.5 | 15' |
| HSSRX-M4112 | 41 | 19 | 12 | 4 | 32 | 3.4 | 10'/15'/30' |
| HSSRX-M4835 | 48 | 35 | 13 | 3.4 | 44 | 3.0 | 30' |
| HSSRX-M5012 | 50 | 35 | 12 | 3.5 | 42 | 3.0 | 30' |
| HSSRX-M5015 | 50 | 34 | 15 | 4 | 45.5 | 3.5 | 30' |
| HSSRX-M5512 | 55 | 40 | 12 | 5 | 47 | 4.6 | 30' |
| HSSRX-M6014 | 60 | 43 | 14 | 3.5 | 51 | 3.5 | 15' |
Two of those columns need a word. H2 is tabulated for every item and explained nowhere, which is worth a question rather than a guess. And PCD is a circle, not a hole pattern; how many holes sit on it is the subject of a later section.
Every drawing carries its own checksum
Four of the images are dimensioned drawings, and each prints a chord dimension between two adjacent mounting holes. That number is not decoration. For a pattern of three holes at 120 degrees, the straight-line distance between neighbouring holes is the pitch circle diameter multiplied by the sine of 60 degrees, or 0.866025.
| Drawing | PCD | PCD x sin 60 | Chord printed on the drawing | Agrees |
|---|---|---|---|---|
| 25.5 x 16.5 x 12 | 21 | 18.187 | 18.19 | yes |
| 34 x 20 x 12 | 27 | 23.383 | 23.38 | yes |
| 28 x 16 x 11 | 22 | 19.053 | 19.05 | yes |
| 30 x 20 x 11 | 24.9 | 21.564 | 21.56 | yes |
All four match to two decimal places. That tells you two things at once.
First, every one of these is a three-hole pattern, not four. The drawings do not say so in words anywhere; the chord is the only evidence, and it is conclusive.
Second, the drawings were built from a real geometric model rather than typed by hand. A copied number would not land on the sine of 60 degrees four times running.
That matters because it settles who to believe when the drawings and the table disagree, which they do.
Three of the four drawings contradict the table
| Item | Field | On the drawing | In the table | Gap |
|---|---|---|---|---|
| M2512 | overall height | 12 | 11.5 | 0.5 mm |
| M3412 | every field checked | 34 / 20 / 12 / 27 / 3.5 | 34 / 20 / 12 / 27 / 3.5 | none |
| M2812 | overall height | 11 | 16 (H1) or 12 (H2) | 5 mm to H1 |
| M2812 | hole diameter | 3.5 | 2.8 | 0.7 mm |
| M3011 | pitch circle | 24.9 | 25 | 0.1 mm |
| M3011 | hole diameter | 3.5 | 3.2 | 0.3 mm |
Only M3412 agrees on everything. The other three disagree on at least one field, and M2812 is the worst: the drawing says 11 mm overall with a 3.5 mm hole, the table says 16 mm and 2.8 mm, and the part number ends in 12.
Three numbers, three different answers, for the height of one part. The code ending in 12 is consistent with neither the drawing nor H1, though it does match H2.
For M3011 the gap is smaller but practical: 3.2 mm against 3.5 mm is the difference between a hole that clears an M3 screw with a little slack and one that clears it by 0.15 mm. If the drawing is right, the table's 3.2 is a press fit on an M3 screw and will not assemble.
Because the drawings verify themselves arithmetically and the table does not, order from the drawing and ask for the drawing of the exact item you are buying. That is a general rule for turned parts, and here there is proof it applies.
A fifth drawing that is not in the table at all
The last drawing in the set is in English rather than the Chinese headers the others carry, and it specifies a body of 90 mm, a pitch circle of 80 mm, a window of 71 mm, holes of 5 mm, an overall height of 21 mm and a flange of 6 mm.
None of those figures appears in the table. The table stops at a 60 mm body, a 51 mm pitch circle, 3.5 mm holes and 14 mm of height. The 90 mm part is 50% larger in diameter than anything listed.
Either the range really does run to 90 mm and the table is a subset, or that drawing belongs to a different product that has been filed with this one. Both happen constantly. You cannot tell from the page, so ask which it is before you assume 60 mm is the ceiling.
Note also that this drawing prints no chord. It is the one sheet in the set that cannot be self-checked.
PCD alone will not let you drill the pattern
The table publishes a pitch circle for every item and never states how many holes sit on it. Those are two different numbers and you need both.
For three holes the chord is PCD times 0.866025. For four holes it is PCD times 0.707107. The ratio between them is fixed, so the error of guessing is the same everywhere: 22.5% of the chord.
| Item | PCD | Three-hole chord | Four-hole chord | If you guess wrong |
|---|---|---|---|---|
| HSSRX-M2010 | 17 | 14.72 | 12.02 | 2.70 mm |
| HSSRX-M2512 | 21 | 18.19 | 14.85 | 3.34 mm |
| HSSRX-M3412 | 27 | 23.38 | 19.09 | 4.29 mm |
| HSSRX-M5015 | 45.5 | 39.40 | 32.17 | 7.23 mm |
| HSSRX-M6014 | 51 | 44.17 | 36.06 | 8.10 mm |
A jig drilled to the wrong one of those is scrap, and on the larger sizes it is scrap by eight millimetres. The drawings answer the question for four items; for the other fourteen you have to ask.
Sensitivity: three settings, and no tolerance on any of them
The drawings are headed with a figure like 30' / 2mm, which is the same convention used across the industry: the angle of tilt needed to move the bubble 2 mm. Converting that to the units you work in gives the table below, and the vial radius follows from the same definition.
| Sensitivity | Degrees | mm per metre | inches per foot | Vial radius |
|---|---|---|---|---|
| 10' | 0.1667 | 2.91 | 0.035 | 687.5 mm |
| 15' | 0.2500 | 4.36 | 0.052 | 458.4 mm |
| 30' | 0.5000 | 8.73 | 0.105 | 229.2 mm |
Read that against the vial series this site also covers. The finest cylindrical vial there bottoms out at 20 to 25 minutes, which is 5.82 to 7.27 mm per metre. The finest setting here is 10 minutes, or 2.91 mm per metre: about 2.3 times more sensitive.
That is a real capability and it is also the source of the biggest omission on the page. There is no tolerance on any of these three figures. A vial that is nominally 10 minutes but made to plus or minus 20% is anywhere between 8 and 12 minutes, and you have no way to know which one arrived.
The vial series publishes plus or minus 6 minutes on its finer items. A cased instrument that quotes a sensitivity twice as fine, and quotes no tolerance at all, is harder to specify than the loose component it is built from. Ask for the band, and ask whether it is measured or moulded.
The finer the setting, the flatter the dome has to be
A vial of radius R spanning a window of diameter D rises by D squared / (8R) from the edge to the middle. That rise is the only physical feature that sets the sensitivity.
| Window diameter | at 10' | at 15' | at 30' |
|---|---|---|---|
| 14 mm | 0.036 mm | 0.054 mm | 0.107 mm |
| 16.5 mm | 0.050 mm | 0.074 mm | 0.149 mm |
| 20 mm | 0.073 mm | 0.109 mm | 0.218 mm |
| 28 mm | 0.143 mm | 0.214 mm | 0.428 mm |
| 35 mm | 0.223 mm | 0.334 mm | 0.668 mm |
| 43 mm | 0.336 mm | 0.504 mm | 1.009 mm |
At 10 minutes in the smallest window, that rise is 0.036 mm. For comparison, the shortest fine cylinder vial in the other series needed 0.079 mm, and a normal moulded acrylic part is held to about 0.1 mm.
So the 10 minute option in a 14 mm window is asking for a dome whose whole rise is a third of a normal moulding tolerance. That is not a moulded feature, it is a ground or a selected one.
There is a useful consequence. Look down the columns: for a fixed sensitivity, a larger window gives a larger rise, and a larger rise is easier to make. A 43 mm window at 10 minutes would need 0.336 mm, which is ten times more forgiving than the 14 mm one. Which makes it interesting that the table offers 10 minutes only on the small and middle sizes and not on the large ones. The limitation there is commercial, not geometric.
Only ten of the eighteen let you pick a sensitivity
Ten items list three sensitivities, meaning the same body is available at 10, 15 or 30 minutes. The other eight list exactly one: M2512, M4013, M4835, M5012, M5015 and M5512 at 30 minutes, and M4014 and M6014 at 15.
This is worth checking before you assume a fine setting is available across the range. If your design needs 15 minutes and a 48 mm body, the answer is that this table does not offer it.
Same body diameter, two different instruments
| Field | HSSRX-M4013 | HSSRX-M4014 |
|---|---|---|
| Body diameter | 40 | 40 |
| Window diameter | 30 | 22 |
| Wall between them | 5.0 | 9.0 |
| Height H1 | 13 | 14 |
| Pitch circle | 34.5 | 34.5 |
| Hole diameter | 3.0 | 4.5 |
| Sensitivity | 30' | 15' |
Both are 40 mm across, both bolt to the same 34.5 mm pitch circle, and they are not interchangeable: an 8 mm difference in window, a 1.5 mm difference in hole, a whole sensitivity step apart.
The 50 mm pair is worse. M5012 and M5015 share a body diameter but sit on different pitch circles, 42 mm and 45.5 mm. Swapping one for the other is not a swap, it is redrilling.
Eighteen items describe only sixteen distinct body diameters, because 40 and 50 each appear twice. Ordering by diameter gets you the wrong one half the time on those two sizes.
Wall thickness varies by a factor of 3.7 and nobody publishes it
Subtract the window from the body and halve it, and you have the radial wall: the material between the edge of the glass and the bolt holes. It runs from 3.0 mm on the 20 mm item to 11.0 mm on the 41 mm one.
That is a 3.7 times spread and it is not a smooth progression with size. Two near-identical bodies make the point: 38 mm with a 28 mm window leaves a 5 mm wall, while 39 mm with a 19 mm window leaves 10 mm. One millimetre more diameter and nine millimetres less glass.
Wall is what carries the mounting load and what decides whether the flange distorts when you tighten the screws. A 3 mm wall on a 20 mm body will pull flat against a curved surface; an 11 mm wall will not. If your mounting face is not flat, the thin ones are more forgiving and the thick ones are more honest.
The seating face is the accuracy you actually get
This is where a bullseye differs from a tube, and it is the thing most often got wrong.
Because the vial is rotationally symmetric, the clearance in the mounting holes does not affect the reading. The part can rotate a few tenths of a millimetre on its screws and still read correctly. What it cannot do is sit on anything.
Any burr, swarf or paint under the flange tilts the whole instrument by the shim thickness divided by the flange diameter.
| Burr or swarf under the flange | Tilt over a 34 mm flange | Share of a 10' budget |
|---|---|---|
| 0.02 mm | 2.02' | 20% |
| 0.05 mm | 5.06' | 51% |
| 0.10 mm | 10.11' | 101% |
So a single 0.1 mm chip under a 34 mm flange uses up the entire 10 minute budget before the part has done anything wrong. On the finest setting this instrument is more sensitive than the surface it is bolted to, which means the seating face has to be machined and cleaned, not just cleared.
That also means the flatness of the seat belongs in the drawing of the machine, not in the datasheet of the level. Worth saying out loud in a procurement meeting.
Checking one on the bench
You do not need a certificate to know what you have bought.
Put the level on a flat plate, pack one side of the flange until the bubble sits centred inside its ring, then add shim under one edge and watch how far the bubble travels. One division is 2 mm of travel, so the shim that produces it is the sensitivity, expressed directly.
On a 100 mm arm, one division is 0.291 mm at 10 minutes, 0.436 mm at 15 and 0.873 mm at 30. Use a 300 mm arm instead and those become 0.873, 1.309 and 2.618 mm, which a feeler stack resolves to about 6% at the finest setting.
Two units from the same carton, tested the same way, tell you whether the batch is consistent. That is the number the page leaves out, and it takes ten minutes to find.
Reading both axes, and what that does not give you
A bullseye tells you which way the bubble leans as well as how far, so it points at the high corner directly. A crossed pair of tubes makes you read two numbers and work out the diagonal yourself, and introduces the alignment error between the two tubes.
What it does not give you is a number. As with the cylinder vials, the ring around the bubble is a target, not a scale. The photographs show some units with concentric rings printed round the bubble and at least one with a full protractor scale on the glass, but the printed scale is not specified for any item in the table and is not something you can rely on receiving.
So treat the reading as inside, on the ring, or outside. If you need a number, the ring spacing has to be on the drawing, and for these items it is not.
Thirteen questions for the quotation
The table plus the drawings are enough to buy from, if you get three things settled first: the hole count, the true height of M2812, and whether that 90 mm part exists.
| Question | Why it decides the order |
|---|---|
| How many mounting holes, and at what angles | PCD alone cannot be drilled; three and four holes differ by 22.5% in chord on every size |
| Is there a 90 mm version, and if not what is that drawing | A drawing with a 90 mm body, 80 mm pitch circle and 5 mm holes is in the image set but absent from the table |
| Which is right for M2812, the drawing at 11 mm or the table at 16 mm | Five millimetres apart, on a part whose code says 12 |
| Is the M2812 hole 3.5 mm or 2.8 mm | The drawing and the table disagree by 0.7 mm, which is the difference between clearing an M3 screw and not |
| What tolerance is held on the sensitivity | Three nominal values are published and no tolerance at all |
| What is H2 | A second height is tabulated for every item and never explained |
| Is the case a turned alloy or a die casting, and what finish | Photos show both natural and black, and the table names no material |
| Is the window glass or acrylic | Scratch behaviour differs, and a scratched window is a lost reference |
| Is the vial fluid dyed, and how does it behave cold | Viscosity sets settling time; no temperature range is published |
| Is the seating face machined flat, and to what | This is the number that actually limits installed accuracy |
| Through holes or threaded inserts | The drawings show plain through holes with no counterbore |
| Is the case sealed against coolant and washdown | No IP rating appears anywhere |
| Which of the two 40 mm and two 50 mm items is the current one | Same body diameter, different window, and for the 50 mm pair a different pitch circle |
Where this sits with the other level pages
The mini cylindrical vial guide does the same conversion for barrel vials, including the tolerance bands, and is the right page if you are buying the loose vial rather than the cased instrument. The PMMA bubble vial guide covers the uncased acrylic bubbles in 5 to 25 mm, which is what goes inside a housing you are building yourself.
Frequently asked questions
Can I drill the mounting holes from the PCD alone?
No. A pitch circle defines where the holes sit but not how many there are. For three holes the chord between neighbours is PCD times 0.866, for four it is PCD times 0.707, and the two differ by 22.5%. On the 51 mm pitch circle that is 8.1 mm of error. The drawings show three holes for four items; the rest need asking.
What does the extra chord dimension on the drawings mean?
It is the straight-line distance between two adjacent holes, and it acts as a checksum. For a three-hole pattern it must equal the pitch circle diameter times the sine of 60 degrees. All four drawings match that to two decimal places, which is how you know the pattern is three holes and that the drawing is geometrically sound.
Why do the drawings and the parameter table disagree?
On three of the four items I could cross-check, at least one field differs. M2812 is the largest gap: 11 mm tall on the drawing against 16 mm in the table, with a 3.5 mm hole against 2.8. Since the drawings verify arithmetically through their chord dimension and the table does not, order from the drawing and request one for the exact item.
Is there a 90 mm version?
One drawing in the set shows a 90 mm body, 80 mm pitch circle, 71 mm window, 5 mm holes, 21 mm height and a 6 mm flange. None of that is in the table, which stops at 60 mm. Either the range extends further than listed or that sheet belongs to another product. Ask before treating 60 mm as the maximum.
What is 30 minutes divided by 2 mm?
That is the sensitivity convention: the angle needed to move the bubble 2 mm. Thirty minutes is half a degree, which is 8.73 mm per metre. Ten minutes is 2.91 mm per metre, with a vial radius of 687.5 mm. The radius follows from dividing 2 mm by the angle in radians.
Is a finer sensitivity always better?
No. The finer the setting, the flatter the dome, and the smaller the window the less rise there is to form. Ten minutes in a 14 mm window needs a rise of only 0.036 mm, which is a third of normal moulding tolerance. It also makes the instrument more sensitive than most surfaces it will be bolted to, so a 0.1 mm burr under the flange can consume the entire budget.
How accurate is it once installed?
Limited by the seating face, not by the vial. Any burr under the flange tilts the instrument by its thickness divided by the flange diameter, so 0.05 mm under a 34 mm flange is 5.06 minutes, or half of a 10 minute budget. Because the vial is rotationally symmetric, clearance in the mounting holes does not affect the reading at all.
Key specifications
Frequently asked questions
- Can I drill the mounting holes from the PCD alone?
- No. A pitch circle defines where the holes sit but not how many there are. For three holes the chord between neighbours is PCD times 0.866, for four it is PCD times 0.707, and the two differ by 22.5%. On the 51 mm pitch circle that is 8.1 mm of error. The drawings show three holes for four items; the rest need asking.
- What does the extra chord dimension on the drawings mean?
- It is the straight-line distance between two adjacent holes, and it acts as a checksum. For a three-hole pattern it must equal the pitch circle diameter times the sine of 60 degrees. All four drawings match that to two decimal places, which is how you know the pattern is three holes and that the drawing is geometrically sound.
- Why do the drawings and the parameter table disagree?
- On three of the four items I could cross-check, at least one field differs. M2812 is the largest gap: 11 mm tall on the drawing against 16 mm in the table, with a 3.5 mm hole against 2.8. Since the drawings verify arithmetically through their chord dimension and the table does not, order from the drawing and request one for the exact item.
- Is there a 90 mm version?
- One drawing in the set shows a 90 mm body, 80 mm pitch circle, 71 mm window, 5 mm holes, 21 mm height and a 6 mm flange. None of that is in the table, which stops at 60 mm. Either the range extends further than listed or that sheet belongs to another product. Ask before treating 60 mm as the maximum.
- What is 30 minutes divided by 2 mm?
- That is the sensitivity convention: the angle needed to move the bubble 2 mm. Thirty minutes is half a degree, which is 8.73 mm per metre. Ten minutes is 2.91 mm per metre, with a vial radius of 687.5 mm. The radius follows from dividing 2 mm by the angle in radians.
- Is a finer sensitivity always better?
- No. The finer the setting, the flatter the dome, and the smaller the window the less rise there is to form. Ten minutes in a 14 mm window needs a rise of only 0.036 mm, which is a third of normal moulding tolerance. It also makes the instrument more sensitive than most surfaces it will be bolted to, so a 0.1 mm burr under the flange can consume the entire budget.
- How accurate is it once installed?
- Limited by the seating face, not by the vial. Any burr under the flange tilts the instrument by its thickness divided by the flange diameter, so 0.05 mm under a 34 mm flange is 5.06 minutes, or half of a 10 minute budget. Because the vial is rotationally symmetric, clearance in the mounting holes does not affect the reading at all.