Water SG less than 1.000

[peteru]

This one has me scratching my head. I have a hydrometer calibrated for 20C, one refractometer calibrated for 20C and another refractometer with ATC. A few months ago I checked the hydrometer with filtered tap water at 20C and it was reading bang on at 1.000

Today I was playing around with the refractometers and put a drop of the filtered tap water on the ATC model. It was reading less than 0%. I checked the temperature of the water and cooled it slightly to 20C, then used the water sample with the hydrometer and it was reading 0.998. The ambient temperature was 22C and the refractometer calibrated to 20C was also reading less than 0% Brix.

Why would that be? Why would all three instruments tell me that the SG of my filtered tap water is not 1.000? In fact, all three actually agree in the magnitude, since the refractometers were reading about -0.5% Brix

 

[MHB]

thermometer is out
free alcohol being delivered through municipal water supply
your still pissed from last night

Pick one
Mark

 

[JDW81]

Also need to bear in mind that the equipment most home brewers use is certainly not 100% accurate. For a high spec hydrometer/refractometer you're looking at a fairly substantial cost.

I wouldn't be too concerned, 0.002 is bugger all to be out.

Do you calibrate your refractometer each time you use it? I notice mind needs to be re-calibrated between brews to ensure it is "accurate".

JD

 

[Lyrebird_Cycles]

free alcohol being delivered through municipal water supply

Would increase the RI, not decrease it.

To the OP: was this in Melbourne yesterday? My digital refrac read -0.2 oP on Melbourne water for a brief period yesterday, then corrected itself. I have no idea why.

 

[peteru]

I was not under the influence of alcohol. I had one pint of beer about 4 hours before doing this. This was in Sydney. My water comes from Ryde.

I'm reasonably confident in the thermometer being right. The hydrometer is a cheapie that came with the Cooper's kit about 16 years ago. Both refractometers are really old, more than 20 years, perhaps even more than 30 years old. I don't have instructions for them and they do not have any obvious means of calibration. One is an Atago 0-32% Brix, the other American Optical AO CAT.10423

Most likely explanation would be "something" in the water. Whatever that something is, it seems to have affected the refractive index as well as the density. It could be a coincidental error in all three instruments, but that's pretty unlikely. I'll pull out the mercury lab thermometer and redo the test later tonight, in case my digital thermometer is out.

 

[MHB]

Well that's why I suggested the thermometer was out, what are the odds that 3 instruments would all go out in the same direction at the same time, verses the odds that 1 (the thermometer) is off?

Same for something in the water, what could it be? any salt would move all three readings the other way, unless it was a water soluble liquid like acetone or dichloromethane, which I'm pretty sure you would notice (no I haven't checked their RI's just a guess).

Probability and Ockham both favor the simplest explanation - thermometer being off.
Which is why I have been saying for years that we should all have one good lab thermometer that we can use as a benchmark, which clearly you do, time to use it.
Mark

 

[peteru]

MHB, totally agree with everything you say. The one thing that is strange is that the ATC refractometer was also out, despite the fact that the ambient temperature was well within it's range.

I guess the other thing to mention is that I was taking the readings indoors under a fluorescent light with an 85 CRI. Clearly that would not impact the hydrometer, but it may have an effect on the refractometers. I should be in a position to retest in another couple of hours.

 

[peteru]

Right, so I came back to this and I am none the wiser.

Pulled out the mercury lab thermometer, the Aldi digital thermometer and the Maverick dual probe thermometer. They all agreed on the ambient temperature, 23C. I cooled down plain tap water and filtered water to 20C. Again, all my thermometers agreed that both samples was indeed at 20C.

Tonight, the hydrometer was reading bang on 1.000 with both the tap water and the filtered water. Yay!

Both refractometers were reading a little under 0% Brix. The difference was a lot less than yesterday. Yesterday I was seeing about -0.5% Brix, today it was more like -0.1% or -0.2% Brix. Neither refractometer has graduations under 0%, so this is just an estimate. Given that there appears to be no obvious way of calibrating either of these dinosaur refractometers and that the error is so small, I'm not going to be too fussed. It's good enough to figure out whether the brew is on track.

However, as I was taking the readings, I also experimented with various light sources around the house to see how much of a difference they make. Wow! It varies from being unable to get a reading to being clearly visible. I'll have to try in daylight too. I have a mix of light sources, LEDs and CCFLs mainly, but also a few incandescents and 85CRI fluorescent tubes. The LEDs (ranging from 3000K to 6500K) were by far the worst light source. CCFLs (both cool and warm, not sure about the exact colour temperature) were better but not great. The 85CRI fluoro tubes were by far the best. The incandescent bulb (in an old hand held torch with weak batteries) wasn't bright enough to take a reading and the other incandescents I have are coloured party lights, so I didn't bother.

Based on what I learned tonight about the difference that the light source makes, I'd either be buying a digital refractometer or at least a refractometer with a built in light source.

BTW: The Atago has an easy to read graticule, but the shadow line is very indistinct. The AO has a much clearer shadow line, but the rest of the optics make it very hard to read and the cover plate is annoying. They are solid instruments, but ancient. Not sure if I should be upgrading or just make do with what I have. Are modern (cheap) refractometers much better than these old ones? I'm kind of thinking that if I'm going to spend more $ on toys, then perhaps a pH meter might be a better use of money.

 

[peteru]

Doh! Found a way to adjust both refractometers. In both instances, there were tiny screws that were covered with a white epoxy. I used a small screwdriver to dig out the epoxy on both and then I was able to reset the 0% line to match up with the shadow line cast by plain water.

Both refractometers were inherited from the same person who used them for a very specific task. Perhaps they were both adjusted to give the correct readings for that particular application.

All good now.

Although, I still have no idea why the hydrometer was reading less than 1.000 yesterday, but was fine today.

 

[MHB]

Glad you found the adjuster, it will make life easier.
Was thinking about the offset and the thought that came to me was that if they were both old and reading about the same offset, that they may have been calibrated at a different temperature. 20oC is the standard now, ones upon a time it was very common to do lab work and calibration at around 15oC.
Sort of makes sense of them both being out the same amount. still cant think why the hydro would have been out, other than the possibility that your digital thermometer was having a brain fart.
Mark

 

[Mardoo]

Ideally use distilled water when you adjust your zero point. However, I'm uncertain that the refractometer see usually have access to are accurate enough for that to make a difference.

 

[Lyrebird_Cycles]

However, as I was taking the readings, I also experimented with various light sources around the house to see how much of a difference they make. Wow! It varies from being unable to get a reading to being clearly visible. I'll have to try in daylight too. I have a mix of light sources, LEDs and CCFLs mainly, but also a few incandescents and 85CRI fluorescent tubes. The LEDs (ranging from 3000K to 6500K) were by far the worst light source. CCFLs (both cool and warm, not sure about the exact colour temperature) were better but not great. The 85CRI fluoro tubes were by far the best. The incandescent bulb (in an old hand held torch with weak batteries) wasn't bright enough to take a reading and the other incandescents I have are coloured party lights, so I didn't bother.

The best cheap lightsource for a refrac is an LED, it's just not a white one.

White LEDs are blue LEDs with a yellow fluorescent chip in front of the emitter so they give two strong peaks at different wavelengths, hence the blurring.

Get hold of a yellow LED with a peak near 589 nm and you'll get very sharp lines. Don't forget to current limit the drive circuit.

Refractometers are designed to use the sodium d line emission at 589.3 nm, so a sodium vapour lamp is the best source but they're much more expensive.

 

[Lyrebird_Cycles]

Ideally use distilled water when you adjust your zero point. However, I'm uncertain that the refractometer see usually have access to are accurate enough for that to make a difference.

They're not precise* enough for it to make a difference (unless your local water supply is really ****).

On standard Melbourne water with a TDS of around 80 ppm** assuming the ions are roughly half and half monovalent and divalent***, the refractive index shift is about 0.000014, equivalent to a little under 0.01oP. A research grade Abbe refractometer**** is precise enough for that to matter, no commercial refrac is.

* Accuracy and precision are different things. Only people like me care.

** Measured today at my kitchen tap in Willy.

*** The official typical analysis from Melbourne water is very close to this and it simplifies the computation.

**** If you want one of these, I found a Bellingham + Stanley in a second hand shop for about $250 the other day. That's overpriced but they were about ten times that when new.

 

[peteru]

I found a Bellingham + Stanley in a second hand shop for about $250 the other day. That's overpriced but they were about ten times that when new.

No thanks. I think I'll be fine with my 30+ year old AO ATC and Atago now that I found where to tweak them with the tiny screwdriver.

 

[Busboy]

Are any of these instruments affected by atmospheric pressure? Would that account for changes in readings?

 

[MHB]

No - well maybe the hydrometer but I doubt there is anywhere in Oz that would be tall enough to make any difference we could read.
Mark

 

[Lyrebird_Cycles]

No, water is pretty well incompressible. Bulk modulus is about 3 GPa so a 1% change would require a pressure of around 30 MPa, about 300 bar

 

[sp0rk]

The best cheap lightsource for a refrac is an LED, it's just not a white one.

White LEDs are blue LEDs with a yellow fluorescent chip in front of the emitter so they give two strong peaks at different wavelengths, hence the blurring.

Get hold of a yellow LED with a peak near 589 nm and you'll get very sharp lines. Don't forget to current limit the drive circuit.

Refractometers are designed to use the sodium d line emission at 589.3 nm, so a sodium vapour lamp is the best source but they're much more expensive.

RS Components is silly expensive (top quality parts tho)
I'll be buying my next lot of bits for tinkering from either Little Bird or Core Electronics
http://core-electronics.com.au/t1-3mm-yellow-led-with-yellow-diffused-lens.html
https://littlebirdelectronics.com.au/products/t1-3mm-yellow-led-with-yellow-diffused-lens

 

[Lyrebird_Cycles]

They used to be terribly expensive but they're not bad now. I have wholesale accounts* with DigiKey and Mouser in the US, they're amongst the cheapest suppliers in the world for name brand components and I often find that RS is cheaper when shipping is factored in. Their service is also very good: I've ordered parts at 1800 and had them turn up at 0900 the next day, to a brewery in a small town in country Victoria which meant we got the brewhouse back on line and were only about 3 hours late to mash in.

I will only buy components from suppliers who give a reference to the original manufacturer's data sheet. I would not, for instance, buy the LEDs you listed at any price: neither site gives a manufacturer's data sheet, so I just assume they are cheap crap.

I can't afford to use crap components. I was burnt badly once because an Australian agency supplied off-brand parts which failed in service so I had to repair a whole raft of devices under warranty which completely destroyed my profit margin.

The lack of a data sheet means I don't know what the dominant wavelength is, what the spectral bandwidth is, what the light output vs current curve looks like, what the maximal current values to which I can design if I should choose to use a square wave drive etc etc.




* I used to run a business designing and building specialist analogue electronics.