Weird Refractometer. What's causing repeat measurements on same sample, to keep going up drastically?

[hungupdown]

Usually, my digital refractometer readings make sense.
But at the moment, successive/repeat measurements taken, on the same sample(s), always keep going up. Maybe from brix 17% to 22% over around 3 minutes. This is with samples from a fermenting, high gravity/ABV imperial stout. Present: OG 1143; corrected SG 1026; ABV 14.5%

Normally after putting a sample on the refractometer, I leave it a minute or two before taking a reading. This allows the sample to reach the same room temperature as the refractometer (normally 17-19°C here, this value is used for temperature correction).
Correction needed is larger when sample temperature is much higher than room temperature (eg during mash, it would be around +3.1). Successive/repeat readings should rise slightly, as a sample drops to room temperature.

But at my fermentation temperature (currently 23°C), the correction to Brix @20°C is quite small, eg brix 20 @23°C is brix 20.2 @20°C. So an immediate reading, while still 3°C too hot, would only be marginally low (correction needed +0.2).

Another factor, can be evaporation while cooling, (greater with a hot sample), increasing the density.

But I have readings going up, by way more than any temperature correction! And it's too cool for much evaporation. I've closed my refractometer sample cover to prevent evaporation anyway (to no avail).

Using the reading taken, immediately after putting the sample in, seems to give the most accurate result. With corrected SG then being a close match to a hydrometer reading.

I've already tried cleaning the prism surface, as there looked to be some dry deposit (sugars?) around the edge.
Stirring the sample, seems to be the only thing that brings readings back down (but never quite as low, as the initial one).
After emptying sample, and checking cal with water, readings then stay consistant.

Could the effect be from fermentation gas bubbles, adhering to the prism.

 

[IslandLizard]

Using the reading taken, immediately after putting the sample in, seems to be the most accurate. With corrected SG then being a close match to a hydrometer reading.

That's what I would go by, hydrometer readings should be trusted when taken correctly.

Stirring the sample, seems to be the only thing that brings readings back down

How the heck you do you "stir" that sample? It's only a couple drops.

(but never quite as low, as the initial one)

That seems to point to evaporation and/or old sugar deposits being redissolved.

 

[tracer bullet]

Hmm, with digital it could be the gage itself. I do know that with my standard refractometer that the sample itself will vary over time, not just temp but if it has a chance to settle. A bunch of particulates in it will skew the readings. I've learned to let my samples sit for a few minutes, at least, and afterwards I get very consistent results. But with the proteins or anything else stirred up they'll be all over the place.

Agree to the above there could be some evaporation going on with a small sample. Consider the settling thing too perhaps?

 

[IslandLizard]

A bunch of particulates in it will skew the readings. I've learned to let my samples sit for a few minutes, at least

Excellent point!

 

[hungupdown]

How the heck you do you "stir" that sample? It's only a couple drops.

First time using the tip of the sampling pipette, then with my little finger.

 

[hungupdown]

Agree to the above there could be some evaporation going on with a small sample. Consider the settling thing too perhaps?

Evaporation needed for brix to increase from17 to 22% would be around 25%.
It's hard to imagine much evaporation happening over just a short time, with a sample that's near room temperature (and sample cover either closed or open), when readings for mash samples, at much higher temperatures, barely change (with the cover open).

My refractometer, can give an updated reading around every 10 seconds, and reading was going up every time. I just stopped taking readings, after increase exceeded 5%.

Three days after last sugar addition, with few surface bubbles. Now initial brix reading 17.8%, dropping to 17.5% after 10 seconds, then steady. Ferment temp 22°C, room 21°C.