# TTB stopped our production due to high proof

## Recommended Posts

One follow up point/question.   While I certainly get the significance of this proof variance in accuracy, your example above assumes an absolute difference.  In other words, you show an absolute increase in proof for the higher-accuracy density meters.  I think, unless I am missing something here, the actual result would be plus-or-minus accuracy that, over a larger sampling set, would potentially result in less average variance.

In addition, the IRS-approved hydrometer and thermometer have accuracy variance.   For example, +- .2 proof for the hydrometer, and +-.75 C for the thermometer.  So we would need to calculate the variance for both for a valid comparison... and also include the variance for all the variables of human error and other unknown influences.

If you review carefully the TTB glass hydrometer method and old standard practice, you will note that the accuracy, precision, and repeatability for a glass hydrometer is better than the nominal accuracy. For example, typical IRS glass hydrometer has 0.2% divisions and 0.2% "accuracy", but in fact the method for doing the readings can give you greater precision than the 0.2% divisions, because a skilled user can read to about 1/4 of a division, or 0.05%, and then you are to make multiple (at least 3) readings, and average them. Hey, that is how it used to be done. The result is that a better than 0.05% accuracy is attainable. These analog devices are highly repeatable, in fact, any variance due to repeatability is primarily due to human error, or skill of the operator. Finally, while nominally 0.2% accurate, if calibrated, they can be calibrated to the best attainable accuracy, or 0.05%. That means +-0.1 proof, within the 0.3 proof needed by the TTB and better than a DMA 1001. Okay, in practice today, most users lack the skill and patience to make this measurement that carefully. But the method is assuming you are a chemist in 1933. By the way, the key to all this is having everything fully equilibrated at one temperature, and a very accurate measure of temperature, and correction from 60F using tables.

##### Share on other sites

On 9/18/2018 at 2:25 PM, Patio29Dadio said:

If it were me as the king of the TTB, I would approve more of these less-expensive meters and just require that they are certified every year along with an annual certified lab test of each proofed product.

Better yet is if Anton Paar or Rudolf would make a cheaper instrument that just met the requirements. They could, the difference between versions is to some degree just software. The DMA 4500 is probably overkill. But they have an inherent interest in having a large price point differential above and below the TTB requirement. One day, a vendor will do so, and everyone will do so.

##### Share on other sites

Sometimes these differences are driven by the measured tolerances of the physical measuring components as they come out of manufacturing.  While the specific part may look identical, through testing some units meet tighter tolerance ranges than others.  Instead of scrapping manufactured units, they are put to use in lower-cost devices.  I don't know that this is the case with Anton Paar, but I've seen this kind of thing done elsewhere, especially when the unit manufacturing costs are very high, and either scrapping or fixing reliability in software is not possible.

##### Share on other sites
On 9/19/2018 at 9:09 AM, Silk City Distillers said:

Sometimes these differences are driven by the measured tolerances of the physical measuring components as they come out of manufacturing.  While the specific part may look identical, through testing some units meet tighter tolerance ranges than others.  Instead of scrapping manufactured units, they are put to use in lower-cost devices.  I don't know that this is the case with Anton Paar, but I've seen this kind of thing done elsewhere, especially when the unit manufacturing costs are very high, and either scrapping or fixing reliability in software is not possible.

Most of the difference has to do with temperature control, equilibration, and correction. This is especially true for the handheld units, which have no temperature control. The quartz oscillation tube measurement itself can be done to very high accuracy even with an inexpensive handheld unit. But you have to have the temperature fixed in the device and measured accurately, as well, and then calibrate the whole thing.

## Create an account

Register a new account