Refractometer Calibration

[Woeski]

Why olive oil?

Confused why olive oil is the substance of choice here.
Yes it is cheap, but will add to the inaccuracy of the refractometer because it is a variable substance.
Surely a better substance is Glycerol. Has a refractive index of 1.4729 at 20C, that does not have variability.
This maps to 25.5% water content.
Yes, it's about £2 for a small bottle at boots, but at least it will work properly. No point have the refractometer if you want to approximate.

 

[Hombre]

I suppose that your's is a well considered answer, given that it comes fourteen months after the last post on the subject

I believe you.

 

[itma]

Well, I for one do appreciate it when someone comes across something to add to a discussion and ... does exactly that - ADD it to the discussion, rather than just starting a new thread, disconnected from all previous discussion. (Which Woeski actually also did, just a few days ago.)

As it is, this subject already has a few threads.

Olive Oil is a bad choice for a real "calibration".
However, those advocating using it for sanity checking or for restoring factory setting do muddy the water by referring to those actions as "calibration" (which they are not).


To take the point made by Hivemaker and o90o, you can live with an error, but only after you know its there and know what it is.
*Knowing* those things, as opposed to presuming them or allowing for the error hopefully being within some arbitrarily set limits, does require a real calibration process.

And since a pharmaceutical grade product is going to be much more consistent than an agricultural product (coming from different crops, countries, processes, etc), it is going to be a much better choice as a reference for true calibration -- checking and possibly even improving the accuracy of the factory setting. Or even for making an estimate of the error in that factory setting.

Unless you do check your instrument against a standard that is 'accurately-known' (strictly, that would be traceable to international standards), you simply don't *know* what the error in your instrument might happen to be.
When o90o says "Just like a car speedo- what it reads and what the real speed is could be up to a 10% error" he means that that is the allowance he makes, based on his appetite for risk. The strict logic of his statement is the assertion that no car speedometer has an error greater than 10% -- and that it is an assertion that I do not accept.
10% is a reasonable estimate for the error, but it is possible it could be more.
Until you check it, you simply do not know how big the error actually is.
And you can check it, either with GPS (which I would regard as a sanity check) or by timing against motorway "mile-posts" which are accurately placed at 100 metre intervals ... Calibration is possible - you don't have to accept an arbitrary estimation of the likely error.


I think Glycerol should be a better calibration standard than olive oil.
I don't know enough about its stability to give an opinion on whether it would be as good as olive oil for checking the consistency of the instrument over a period of years -- but there should be much better consistency between fresh Glycerol samples, so getting a fresh bottle would overcome worries about stability.
Similarly, I don't know how Glycerol would compare against purpose-specific optical calibration oils. But I suspect it would be much cheaper and more readily available.

 

[oliver90owner]

The strict logic of his statement is the assertion that no car speedometer has an error greater than 10%

The strict logic of my statement was the law (as it used to be in the olde days) - a speedo is defined as legal as long as it performs within those limits. If your speedo is inaccurate by more than tem percent you are strictly outside the laws of the road. If the speedo is within that range of accuracy you are legal, but that does not allow you to exceed the actual speed limit in force ie you need to know the error!

Electronic speedos are generally far better than the old mechanically driven ones, which could not be 'converted' when fitting different sized tyres, for instance.

 

[itma]

My point was not about motoring law.
It was about discovering the extent to which the reading was in error.

If you start by assuming the speedo is legal, and no more, then you are assuming the error could be up to 10%.
But you are still only assuming.

 

[oliver90owner]

Motorway markers are exceedingly accurate. I happen to know that the 607 is as spot on as can be expected; I know the saxo is 10% too fast at 30mph and about the same error at 70.

The Saxo error is likely down to lower profile tyres and an antiquated speedo drive system. Most speedos are optimistic if they have an error - better that way.

Tachometers were far more reliable and we used to check them against the mains frequency. That was in the good olde days.

A refractometer is a breeze to operate and use. It would not be too difficult to work out the limits closely, but why try to sail close to the wind and relying on one reading or even one type of measurement. It is just another tool, that is all. I rarely ever need to use mine.

 

[itma]

... I happen to know that the 607 is as spot on as can be expected; I know the saxo is 10% too fast at 30mph and about the same error at 70.
...

But unless you calibrate (or even check) your refractometer against a known well-fixed standard, as you have done with your speedometers, you can't know how much error there might be in your refractometer readout.