Any thoughts on where to find "true level"?

[Spark plug]

Get a tray of water big enough to fit the level in, put it in the freezer, wait for it to freeze, then without moving the tray you should have a flat surface.

 

[Sean]

put a longish spirit level on a horizontal ish surface shim it up to level with paper etc, flip it 180 to make sure its still level , put your gizmo on top of the level and re set.

* flip = turn end to end

 

[Parm]

Option 1 - DIY http://imextools.co.uk/Digital_Level_Calibration.pdf

Option 2 - DIY.

Option 3 - DIY.

Option 4 - fully traceable UKAS to primary standards. The Dogs Danglies in precision and accuracy, but probably cheaper to buy a new level

https://www.npl.co.uk/calibration

 

[rikrobson]

So the easy way is to get a wooden batton and fix a tube to it and fill with water. The water will level itself. Then you have a reference. Get that level and you are done.

If you are up in Perth I have an engineers level, surface plate, some bars and gauge blocks we can sort it out.

 

[500e]

And I thought I was not easily confused

 

[DAPPH]

When I had a decent ( ish ) laser level I used a 400 x 400 mm sheet of thick glass with three balls of blu tack in a big triangle under it set on the work top in the kitchen as it is a fairly solid base to work from . Set the plate by gently pushing one side checking it then set the opposite side at 90 degrees to the side you have just set so you get zero .

Now check all four sides & the diagonals in both directions by rotating the level 180 degrees each time , keeping its base still on the glass . If all these readings show level enter it in as 0 or level & lock it if you can .

 

[Screwdriver]

So the easy way is to get a wooden batton and fix a tube to it and fill with water. The water will level itself. Then you have a reference. Get that level and you are done.

If you are up in Perth I have an engineers level, surface plate, some bars and gauge blocks we can sort it out.

LOL. So basically, you’re suggesting he should make another spirit level!

Personally I don’t think he deserves one...

 

[skotl]

LOL. So basically, you’re suggesting he should make another spirit level!

Personally I don’t think he deserves one...

Can't find the middle finger emoji

I think we're getting somewhere! If you set it on a "known" 30 degree slope then accidentally hit calibrate, it will call that angle zero. Are you with me?

Now when you turn it through 180 degrees and put it back on the same slope, it will read 60 degrees (or 300 depending on your model of level). DO NOT TURN IT UPSIDE DOWN!

So let's pretend you can stuff a wedge under that 30 degree thing, you want to wedge it by half the error being displayed which in this case we happen to know is half of 60 i.e. the 30 degrees you want to remove to give yourself a calibrated level surface. In general though, even if you don't know what the error angle is when it is reading zero, just turn it 180 and wedge up the surface until you remove half the reading, then hit calibrate, then turn it again, and again, and again...

Yes! Now I'm getting it! That does make sense now - thanks for persevering

 

[ukracer]

By jove, I think you've got it! Probably best not to use a scribe or permanent marker, but that definitely sounds like a plan

Just take it back to shop... Get a replacement.

 

[DAPPH]

Slips of cut greaseproof paper make handy shims when sorting out level to a great accuracy

 

[mike os]

It will have a recalibration method, usually needs a reasonably level surface.

Calibration is 2 or 4 stage... set it on level, calibrate 1, turn it 180 degree calibration 2. It should work out itself where it needs setting. Exact procedures is brand specific

 

[gaz1]

if i was doing it id choose a water level

I could say id choose a laser level but i know that can go out and so can a bubble level

cheapest laser level

Accurate to 0.5mm/m

the 100.00 level

Measuring Accuracy 0.5-0.75 mm/m

the cheapest option of all is water

and that is 1/16th off level

http://www.factsfacts.com/MyHomeRepair/WaterLevel.htm

an engineers level

Sensitivity options 0.1, 0.05, 0.02 and 0.01mm/m

oh and eye leveling isnt always true

 

[Screwdriver]

I can't believe this thread is still going or that people are still suggesting you introduce a secondary level to measure it against.

Using water to measure a "level" is all well and good but you then have to transfer that measurement to calibrate the equipment under review. That will lead to inaccuracy. You can't use a flat water plane to support a spirit level or laser level during calibration (freezing which I think** was a joke, doesn't work because water does not expand uniformly as it freezes) so you would have to use that setup to calibrate a surface by eye. Water level in a tube forms a meniscus which deforms the surface meaning you have to average it out by eye. Fine when you're building a house and measuring greater distances but at smaller scales (remembering you are setting up a surface) the errors introduced when the "level" is read off can be huge. That is why a "spirit" level is so-called, they use spirits, typically a very thin oil which does not form a meniscus or suffer from friction at the tube interface.

Even if you did manage to create a "perfectly" flat surface, you still have to place the unit under test on it and it still might not register a perfect level due to inaccuracies in the unit itself. Those built in errors still need to be dialled out by taking two measurements and adjusting the unit itself until it gives you a zero in both orientations. Typically with a lit. sprit level, you mechanically adjust the bubble chamber with respect to the body of the unit and specifically wrt the base. An electronic device will perform that adjustment by introducing an offset in the readout.

With both sprit based and electronic detectors, the theoretical precision of the measurement process is much higher than any water based system, otherwise why not just use water in the first place.

A spirit level or laser level is self calibrating. The most accurate way to set the level is to use it to measure its own reading until it registers a zero in both orientations along a line set up on a surface as described above (many times now). It will then give you a reading which corresponds to the ultimate accuracy of that particular level.

Can we move on to "set squares" now? They are another self calibrating measurement system....

 

[Keith 66]

I used to use a water level often in the boatshop for setting boats up for build or restoration, long bit of clear pipe, water & some food dye if you were feeling rich. Over thirty feet it was as accurate as i needed it.

 

[Climber]

I used to use a water level often in the boatshop for setting boats up for build or restoration, long bit of clear pipe, water & some food dye if you were feeling rich. Over thirty feet it was as accurate as i needed it.

That's exactly the point, it will be very very accurate over large distances. No good for very short distances though.

 

[Screwdriver]

I used to use a water level often in the boatshop for setting boats up for build or restoration, long bit of clear pipe, water & some food dye if you were feeling rich. Over thirty feet it was as accurate as i needed it.

Point being, yes you've got two perfectly level dinky little puddles of water thirty feet apart but what do you do next? By eye you'll scribe a line as close to the surface as possible and in that instant introduce a surprisingly large number of errors which, on a boat will be neither here nor there but will be orders of magnitude less accurate than say, a machine level.

For house building, boat building etc. the water tube level is a fantastic tool but a dedicated levelling instrument should be based on a much more accurate sensor and since they are inherently self calibrating (I think I've said that before) you're always going to be better off using the instrument itself to generate a level surface for calibration.

 

[skotl]

I can't believe this thread is still going or that people are still suggesting you introduce a secondary level to measure it against...

Well, I've learned loads from this thread so I'm happy

 

[Seadog]

That's exactly the point, it will be very very accurate over large distances. No good for very short distances though.

I think you misunderstand what Keith wrote. My reading of it is that at or around 30 feet, the size of vessel he dealt with, it was accurate enough. Why should it not be accurate at smaller distances?

 

[gaz1]

Well, I've learned loads from this thread so I'm happy

Its also answered one of my own questions