Sciencemadness Discussion Board

mercury thermometers

beastmaster - 12-4-2010 at 14:18

I have been trying to buy some mercury thermometers from several places I have always dealt with and they either no longer carry mercury thermometers or they don't ship to the state I live in
anymore. I need some thermometers in the 200C range.
This has something to do with many states blocking products that contain mercury from entering the state. Man what a pain. I have found some sources for really expansive ones, but I have always bought cheap ones around 10.00 a piece.( i often break them) Non mercury thermometers don't work well at higher temps I am told. Has any one experience with an alternative method to measure temps during reactions and distillations? Such as Thermo couplers or the non-contact types? Thanks for any feed back. beast.

thereelstory1 - 12-4-2010 at 18:59

Beast,

try the ir therms you can get. Is this practical for your app?

unionised - 13-4-2010 at 11:10

Have you tried the sort of store you would buy cooking equipment?

Sugar thermometers and oven thermometers will reach 200C or so. Above that you might want to look at thermocouple type thermometers.

jim30082 - 13-4-2010 at 12:06

I get mine from Sargent-Welch.

Panache - 16-4-2010 at 05:54

whats wrong with a teflon coated thermocouple, i find these far far superior to mercury thermometers, not that i do not enjoy mercury thermometers, i actually collect them (not actively just if they cross my path), but they are just, well crap on every level c/f the thermocouple.

Wizzard - 16-4-2010 at 11:04

A wax casting alcohol-based themometer goes from 32-450*F.

densest - 16-4-2010 at 11:19

@panache - do you find the thermocouples are consistent from unit to unit? I read at the Omega website that + or - 1 C is normal tolerance. Do they stay consistent?

I've been looking at platinum resistance sensors which are actually quite inexpensive if you buy the bare beads. Figuring out how to get 4 teflon coated leads on an 0805 surface mount resistor and sealing it all in teflon is the challenge!

@wizzard - I've had some problems with alcohol thermometers of that type: 3 units had a spread of 6C in multiple tests, none were within 2C of a calibrated mercury thermometer. The liquid column tended to separate and was very difficult to fix. The response was 30 seconds or more versus 2 seconds or so for the mercury thermometer.

Do you have a brand which works well for you?

bbartlog - 16-4-2010 at 12:09

I had separation in my alcohol thermometer (which did however mysteriously correct itself a few days later). Generally I prefer mercury. The mercury ones that I got were both from United Glass Tech, which obviously is a reseller and not the manufacturer. While they are functional I found that boiling ethanol removed the external markings (I had to re-mark using the faint remaining lines). So I'm still looking for a mercury thermometer with markings that are actually fused into the glass.

'whats wrong with a teflon coated thermocouple'

I should look into it. Expense might be one factor. But maybe the advantages are enough that I'd get one anyway.

beastmaster - 16-4-2010 at 17:57

Do thermocouples come in a size same as a thermometer so it can be used with the same holders and fixtures i was using the thermometer with?(distilling heads and what not) I seem to break glass thermometers faster then I can get them.

watson.fawkes - 18-4-2010 at 07:44

Quote: Originally posted by densest  
I've been looking at platinum resistance sensors which are actually quite inexpensive if you buy the bare beads. Figuring out how to get 4 teflon coated leads on an 0805 surface mount resistor and sealing it all in teflon is the challenge!
My first thought was to use a small carrier circuit "board" made of copper-coated mylar or kapton. You should be able to make one only about 6 or 7 mm wide. Use a double-sided material. Instead of vias, use a couple of U-shaped wrap-around jumpers on the far end of the carrier. Solder two wires on each side of the near end.

Then, instead of sealing in teflon, put it down the end of a sealed 8 mm ID glass tube with a drop of silicone oil as a heat-transfer medium. If you need teflon encapsulation, the best low-end way I've thought of is a slow-operating injection mold rig, when you don't overheat the teflon and treat it as a very viscous fluid. This might not be the best production method, but it would work fine for a few units.

densest - 18-4-2010 at 10:01

@watson.fawkes - yes, all of that could work quite well - the carrier board & molding suggestions are good - I've done less refined things in the past. I wrote a long reply before searching for commercial RTD probes and was astonished. Omega sells them (class B +- 0.25C @0C) for $50-100, which is a lot less expensive than I thought.

What astonished me was this: Auber Instruments http://www.auberins.com/index.php?main_page=product_info&... sells a platinum RTD probe for $16 :o good to 250C! Immersible (stainless steel & PFA) too. Accuracy class A! +- 0.15C @0C. 4mm X 50mm case. Best deal I've seen by far. It would fit in a straight tube which would fit in a straight thermometer-to-ST-joint adapter, so the only fabrication would be to cut off a length and close the end of an appropriate diameter tube.

Other manufacturers sell RTDs rated up to 500C, but thermocouples are quite useful beyond 200C or so and type K works to over 1300C.

I think I might buy one of the Auber probes. I have Omron & other temperature controllers which accept RTD and all common thermocouple inputs. The controllers are pretty expensive new ($60 for cheap Chinese brands, $150-500 for US/European brands) but available used for a lot less. The whole setup is attractive - remote reading & rugged, though it requires power. There are battery powered converter-displays but I'd have to buy one. Even with lead error, self heating error, and conversion error, the result should be within +- 1C! I like this :D

Ephoton - 18-4-2010 at 10:16

I use thermocuples but use a glass tube for vac distillation to encase them.

at the end I wrape the opening with teflon no probs up to a few hundred degrees.

they are a lot more acurate than I thought they would be.

the only problem I have is that the glass takes time to heat up so the thermocuple is a few degrees behind
but you get used to it.

also using a thermocuple and a PID for your mantle works a dream when you get used to it.

densest - 18-4-2010 at 10:50

@ephoton - I think that watson.fawkes suggested putting a small amount of silicone oil in the glass tube to conduct heat from the glass to the thermocouple. That should speed things up a lot. Yes, a PID controller is really a dream to use!

watson.fawkes - 19-4-2010 at 07:51

Quote: Originally posted by densest  
The controllers are pretty expensive new ($60 for cheap Chinese brands, $150-500 for US/European brands) but available used for a lot less.
Auber sells these too. This one is only $42. It takes RTD input such as the type from the original link. It looks like a clone of a Watlow 92, a couple of which I've bought used for not very much less.

BTW, thanks for the initial pointer to Auber. They also have panel mount mini-connectors for K thermocouples, a part I couldn't find cheap last year. (I rewired a thermocouple with a heavy ceramic plug, which is a bit overkill, but I picked up a box of plugs and sockets cheap.)

Panache - 20-4-2010 at 15:07

Quote: Originally posted by Ephoton  


the only problem I have is that the glass takes time to heat up so the thermocuple is a few degrees behind

this is just as true for a mercury thermometers, they also are encased in glass, the drop of Silicon oil in the tube is good, however why a drop of mercury, better thermal conductor, wait i think mercury may be toxic, need to check that (that was a joke)

watson.fawkes - 21-4-2010 at 05:26

Quote: Originally posted by Panache  
this is just as true for a mercury thermometers, they also are encased in glass, the drop of Silicon oil in the tube is good, however why a drop of mercury, better thermal conductor, wait i think mercury may be toxic, need to check that (that was a joke)
The larger problem is that mercury is an electrical conductor, as are all the low-melting alloys like those used in fire sprinklers. It's easier all round to use an electrically insulating fluid.

densest - 10-5-2010 at 16:24

I finally got around to testing an Auber platinum RTD - it looks good! I connected it to an Omron E5CJ temperature controller as a Japanese RTD sensor.

In a quickly stirred ice/water bath it registered 31.9-32.0F.

I put 5 standard thermometers (4 mercury: -10-110C, -20-150C, -20-300C, -20-500C, one alcohol -30-500F) and one electronic (el-cheapo Pyrex brand meat thermometer 0-300C) in a pile with the RTD and let them stabilize for 2 hours. The alcohol thermometer is useless - the column separated and can't be rejoined easily, and even compensating for the bubble it's an outlier by 6F! The mercury thermometers were all within 2 C: the high range ones read 20.5, 21.0 C and the low range ones read 22.5, 23.0 C. The Auber unit read 22.1C, and the electronic one read 21.7C.

It's somewhat more difficult making an isothermal 100C bath - water superheats easily and one can observe as much as 5C between the top and bottom of a water bath unless almost violently stirred. Perhaps later.

The unit seems quite usable. I tried it with a DIN/ISO calibration and it read about 1C high, so the Japanese curve is probably correct.

not_important - 10-5-2010 at 18:30

You might be interested in the old melting point reference standards, several are not difficult to get. Salicylic acid at 159 is another, and there's always the freezing point of high purity indium under an inert gas :-)



Attachment: WHO Melting-Point Reference Substances.pdf (674kB)
This file has been downloaded 611 times

bquirky - 10-5-2010 at 18:50

I purchased a -10 to 350dec C mercury thermometer from here

http://www.wiltronics.com.au/catalogue/shop.php?cid=118511

it cost me $3.75 au. it shipped in a hard plastic tube.

I cant speak for its accuracy. as i have no others to compare it too.

densest - 10-5-2010 at 20:24

@not_important: if I could figure out an easy way to separate indium from lead I'd be happy - I've got 5 kilos of 50% indium solder! Salicylic acid isn't hard to come by - whistle compound ;) Of course vanillin is a very pleasant aroma too!

@bquirky - I've had very good luck with mercury thermometers - all of them I've tried have been pretty accurate (+- 2C).


[Edited on 11-5-2010 by densest]