Scale Calibration

#1: Post by **Mochamaker** » July 1st, 2021, 10:50 am

I've been playing with espresso for 15 years now. I have never calibrated a scale. Should I buy some calibration weights and do it?

Dan

#2: Post by **danetrainer** » July 1st, 2021, 11:42 am

Probably not depending on the scale, it's resolution and how accurate you need the results. You don't say what brand of scale you have, but you can google weights of common items to place on your scale to see if it's close.
One common test is with coinage and a US nickel is designed to weight 5 grams.
I just put one on my Acaia Pyxis scale that I have the resolution set to hundredths of a gram and it measured 4.96, I tried my Lunar that is set to tenths and it displayed 5.0g.

#3: Post by **Jeff** » July 1st, 2021, 12:39 pm

I calibrate all of mine with a graded weight. I often find them off by 0.5 g or so at 500 g. It is mainly for consistency across scales.

I would suggest an M2 or better weight. They aren't very expensive. https://labbalances.net/blogs/blog/guid ... on-weights for details of the various grades.

#4: Post by **BaristaBoy E61** » July 1st, 2021, 1:20 pm

Before purchasing a calibration weight, which I would recommend, first check to see that your scale has a option of calibration. Not all scales have a user calibration mode & option.

#5: Post by **jpender** » July 1st, 2021, 1:30 pm

A scale error of 0.5g at 500g is 0.1%. If you're weighing out 20g of beans that's an error of 0.02g or about 1/7th of a typical coffee bean. That's probably less than the scale precision.

I never calibrate my scales for coffee.

Check it with some new looking nickels. If your scale is far enough out of calibration to matter for weighing coffee doses/shots I would suggest that rather than spending the money on a good calibration weight that you instead spend the money on a decent scale.

#6: Post by **Marcelnl** » July 2nd, 2021, 10:14 am

most scales do not even have the option to calibrate, checking some newish coins should work, weighing a known volume of water works too as that coincidentally weighs 1 g/ml @20'C using enough volume to decrease the error in volume should work.

#7: Post by **RyanJE** » July 2nd, 2021, 10:16 am

Ive tested scales in the past using coins. A nickel should weight about 5 grams. Therefore use as many as you want and it will average out.

#8: Post by **sweaner** » July 2nd, 2021, 10:19 am

A US nickel weighs 5 grams. Just test with some.

#9: Post by **Jeff** » July 2nd, 2021, 1:00 pm

I've found coins in circulation to be worn enough to be good for a quick check, but not calibration. They don't "average out" as they all tend to be under weight.

#10: Post by **jpender** » July 2nd, 2021, 4:04 pm

For sure, they start out at 5g and unless caked with dirt they lose weight with wear and tear. I imagine you can get a handful of new nickels at the bank. Otherwise look for ones that are shiny and have no scratches but still expect the reading to possibly be a little low.

Here are four nickels out of a jar that are well used:

4.961g
4.909
4.920
4.958

4.94 ± 0.03 (average and std dev)

Now here are four nickels out of the same jar that look pretty new but not perfect:

4.974
4.993
4.998
4.960

4.98 ± 0.02 (average and std dev)

The better four add up to 19.93g. If you have a scale that reads to the nearest 0.1g it should report 19.9g. Is that good enough? You have to decide how precisely you want your doses to be measured.