How to buy an accurate thermometer

The question is often raised about the accuracy of thermometers within our tanks. Whilst some prefer the new digital one, some prefer the aged old mercury glass thermometer. But can we trust these little instruments?

I've seen the difference between 2 glass thermometers that are literally 5 degrees apart. So then how can we be certain that we can be almost certain that the reading we are getting is at least within 90% accuracy?

Well, we are all a little looney when we walk into a LPS, and I've heard people comment about "us fishy types"... so why not just reinforce the idea that we ARE in fact a special breed of people, and need to be treated with utmost care.

When purchasing a thermometer, in order to make certain you are getting something accurate, pull ALL the thermometers off the shelf. Yes, that's right, ALL of them. Now place them side by side. Logic tells me that if they were all in the same spot, they should all be giving the same reading? Yeah, RIGHT! The market is flooded with poor quality junk, so we need to make do with what's at our disposal.

Now take a reading off of all the thermometers. Let's say you have 10 thermometers, and they are reading:
1=24deg
2=25deg
3=19deg
4=26deg
5=29deg
6=24deg
7=24deg
8=21deg
9=22deg
10=25deg

If you add up all the degrees you get: 239
Divide this by 10 (number of thermometers) and you get 23.9
Thus you purchase any one of the thermometers that is showing 24 degress

cguru:

Now that, ladies and gentlemen, is how you buy a thermometer!

 

Thanks @Zoom, really appreciate your articles, think they might look at you like from which mental institute did you escape if your packing out all of the thermometers in order to buy one, but as they say there is method in the madness, :nuts:

 

It is possible that the thermometers have different temps. eg the thermometers at the back of the row are likely to have a cooler temp than ones packed in the front of the row.

I'm not sold on using a average to judge it's accuracy. I would much rather just stay clear of alcheapo glass thermometers. If one of them is inaccurate, it is likely that the whole batch is inaccurate.

Just my 2c, use it, dont use it.

 

Ok but this goes for heaters too, how accurate are heaters if you set it to 24 degrees and the thermometer shows 23 or 25 degrees. Will we ever get it perfect - i don't think so

F

 

@Verndog - The problem with inaccuracies is not limited to cheap thermometers. I have 2 of the digital Boyu submersable range and they read 1 to 2 degrees different. I also have a cheapo glass heater which reads 4 degrees lower than those 2. So I would say that @Zoom's method is pretty clever and yes, fairly insane. Not sure my LFS would appreciate me taking all their thermometers off the shelf. :laugh:

 

@Fonkie - Heaters are also unreliable and depending on the size of your heater and tank you could land up with massive swings in temperature as I have experienced lately.

 

I'm sorry to say this too... but those digital Boyu submersible thermometers are considered cheap (and inaccurate) too.

 

@Zoom - Really? I didn't think R60.00 was very cheap considering a glass thermometer is R10.00...What now?

 

I had a boyu digital once and it was gravely inaccurate... besides needing a battery replacement every month.

 

I find the ones with a smaller amount of liquid are more accurate, the longer thiner ones are more accurate then the shorter fatter one.

I assume this is because of that liquid they use inside the thermometer expands according to the temp and having a smaller amount spread over a longer surface is easier to measure.

Oh another thing medical thermometers, glass ones from dischem pretty accurate.

 

@Zoom try your baby's thermometer and say what it says:redface:

 

While you all go on about thermometer accuracy, please bear in mind that Repcal (the place that calibrates digital thermometers in South Africa to NIST standards) calibrates and verifies with a 0.5C accuracy, so two thermometers can literally differ a whole degree and still be labeled as suitable for medical and pharmaceutical use...

Yes, average is not the best way of finding the most accurate thermometer, but in lieu of carrying around a calibrated thermometer to check yours (That's why I do, but I have one available to me), I think average is not that bad. You could use your baby thermometer, but I'm very very sorry to tell you that the error margin on that is 0.5 - 1C... Normally pretty close to real, but it might be out with a degree and still be seen as good.

If any of you feel strongly that other SA services exist which can calibrate thermometers with a smaller error margin, pm me, I would like to know. Obviously it has to be a SANAS accredited lab, otherwise it does not mean much if they can tell you it is exactly 20C in you kettle...

 

But surely the kind manufactured for humans is better controled then the kind made for fishies that dont really complain

 

I'm not so sure my wife will let me use my daughters in ear thermometer. :laugh:

 

ear thermometer will not work as the water is too cold. I do not think they can read below 30.

 

Great Suggestion! I think I am going to use that logic in business... hehehe let you know how it goes.

 

@Zoom your logic only holds if you are the first few to pull the "working" thermometers of the shelf ... hope we do not shop at the same LPS.

The more TASA members that do this to the same set the less likely it is that you will be the purchasing the accurate remainders.

So I suggest a question to the astute LFS owner before you do this ... Dear Sir has anybody done a sampling exercise on your current thermometer stock that involved calculating the mean temperature. If you get an "@!##@$ ... not you again .." then you know your quest could be doomed.

Later Ferdie

 

hmmm interesting topic.. what i do is set the heater to "X" temperature and then later i check to see if the tank thermometer matches that temperature, if its reasonably close, say within a degree or so or maybe even identical, i'll take that as good enough for me.

 

I was talking about alcheapo glass thermometers since we have already concluded that the digital thermometers are highly inaccurate.

Intresting post, I think we all go on about accuracy in most things in this hobby because it's flooded with cheap chinese alternatives. Specifically talking about the cheap glass thermometers in @Zooms artical.

I'll admit I have a micro sample according to standard, but If I put my SERA Percision glass thermometers(3) they all read within a degree of each other. You looking at around R70.00 for one, which is madness but I guess you get what you pay for.

The problem in my mind is that the QC in these cheap glass thermometers is none exsistent. So a average out of the same batch could be far from accurate because the average in the batch is way out.

Just my way of thinking, but each to his own.