Uber High Tech Thermal Probe Calibration Device

[Martinm210]

Special thanks to Virtualrain, this actually works pretty well. I think the pictures speak for themselves, and no....you don't always get what you pay for..:D
http://img175.imageshack.us/img175/7143/thermometerbummereq2.jpg

[dinos22]

you gotta stir the water and let it sit in there for a little while to get zero reading

that's what i asked sellers of thermometers to tell me what the reading's at :D

[Martinm210]

you gotta stir the water and let it sit in there for a little while to get zero reading

that's what i asked sellers of thermometers to tell me what the reading's at :D

Yeah that was stirred...the cheap probes were right at 0, bounced around +- .5C.

The multimeter was stuck way high. It doesn't correlate with any of my other thermometers. Fortunately you can run a "relative" mode and zero it out, but it's only temporary, no permanent means to calibrate it.

[Exahertz]

Do they read 100c in boiling water?? Or is that off too?

My question is, how high of quality are the thermocouples? And what type are they (J or K)? If the two metals are not perfectly bonded and/or if there are impurity’s in the two metals, you will get some discrepancy in the voltage reading between them.

I work with some very nice TC equipment which can calibrate to account for the different voltages of the probe. Using known temperatures (like ice'd and boiling water) I can set a low and high point thus giving a more accurate reading along the scale. Most TC's of the same type are slightly different unless they were made the exact same way and from the exact same wire material.

But it strikes me as odd that those units have not been calibrated very well. Two of the same meters reading .5c difference is WAAAYYYY OFFFFFFF (or their really cheap-o units). As for your multimeter (like my fluke 87, and some of my other "beater-meter's") you can open your meter up and you should find a potentiometer that you can calibrate the temp readout with. Just be sure to check the accuracy of the scale by dunking the probe in some ice water (0c) then in some boiling water (100c)
<Note: for repeatability and accuracy do this on a warm sunny day when the pressure is around 29.9 In.Hg>

[Praz]

Think I'll continue to have Fluke calibrate and certify my meters once a year.

[Martinm210]

do they read 100c in boiling water?? or is that off too

I don't think that works because you're not changing a very defined state like solid to liquid. With ice water it will relatively accurately hold right at 0 degrees because you have both states of water together. The water will melt the ice until it reaches zero and if stirred will reach an equilibrium or close to that as the room slowly warms the water. But because air is an insulator, heating the water is a slow process....should be fairly close to 0 degrees.


Think I'll continue to have Fluke calibrate and certify my meters once a year.

That would be the best way, I just thought this was a cool and cheap check for the average user that just wants a quick check.

[dinos22]

Do they read 100c in boiling water?? Or is that off too..

depends on what altitude you live :D

[Exahertz]

depends on what altitude you live :D

That’s why I mentioned pressure needs to be 29.9 In.Hg. But you can find altitude/pressure conversions and use that pressure value in a pressure/boiling point of water conversion

[Exahertz]

I don't think that works because you're not changing a very defined state like solid to liquid. With ice water it will relatively accurately hold right at 0 degrees because you have both states of water together. The water will melt the ice until it reaches zero and if stirred will reach an equilibrium or close to that as the room slowly warms the water. But because air is an insulator, heating the water is a slow process....should be fairly close to 0 degrees.

its all about "Latent heat" and it works in all three states of matter :yepp:

[Martinm210]

its all about "Latent heat" and it works in all three states of matter :yepp:

You're right, I should try that, just not sure I want to on the plastic probes...:D

[dinos22]

That’s why I mentioned pressure needs to be 29.5 In.Hg. But you can find altitude/pressure conversions and use that pressure value in a pressure/boiling point of water conversion

true that

[Martinm210]

Here is a calculator for the boiling point of water based on altitude:
http://www.csgnetwork.com/h2oboilcalc.html

It should be 95.8 degrees for my 4,200ft elevation.

[Exahertz]

[edit all] sorry i just ran the math and the pressure is 29.9 in.hg for a boiling point of 100c

[Exahertz]

Also, you must use Pure Water like distilled, do not use tap water! If there are even minute traces of salt, chlorine or other minerals they will change the latent heat of the fluid a little.

[MpG]

Wouldn't trace amounts of salt change the actual freezing point too? Or would that only apply going from liquid to solid?

[Martinm210]

Hehe, well I went ahead and tried this anyway. No not distilled water, but filtered water so it's probably off.

Anyhow, it's been a while since my chemistry class days, but I ran a quicky anyhow...:D

http://img112.imageshack.us/img112/1246/thermometerbummerboilindv2.jpg

Not accurate, but interesting. I remember doing a lab a LOOOOONG time ago calculating the amount of heat energy in steam (Huge!), but that was WAY back.

Anyhow, the ice water test is probably more accurate, easier too.

I have a Fluke imitation on the way with dual probes....I hope it does better than this one. I might have to pull this one apart and see if there is any adjustments inside.

[Exahertz]

Wouldn't trace amounts of salt change the actual freezing point too? Or would that only apply going from liquid to solid?

I believe salt move everything up, so your freezing point is higher and your boiling point is higher. But I'm not sure of that.

Hey Martin. While you’re running your test, after everything’s done, could you throw some salt in the boiling water and let us know which way the temps move???

ahh darn i guess its a little to late for that test seeing as you just posted pics and im pretty sure your done cooking thermocouple stew... LOL

[Martinm210]

I believe salt move everything up, so your freezing point is higher and your boiling point is higher. But I'm not sure of that.

Hey Martin. While you’re running your test, after everything’s done, could you throw some salt in the boiling water and let us know which way the temps move???

ahh darn i guess its a little to late for that test... LOL

Salt raises the boiling point and lowers the freezing point.

I remember the salt trick from cooking tips (throw some salt in when boiling noodles to cook them faster), and DOT's (where I work) some salt the roads or other chemicals like Magnesium Chloride (which is what Oregon uses) to lower the freezing point.

So yeah, probably any chemical has the ability to influence the temperature, you'd have to freeze distilled and pour a fresh cup of distilled water to be more accurate.


I can see it now....As you give the wife the evil eye...."Those aren't my "Distilled" ice cubes in your diet coke are they?"...:D

[Exahertz]

Salt raises the boiling point and lowers the freezing point.

I remember the salt trick from cooking tips, and DOT's (where I work) some salt the roads or other chemicals like Magnesium Chloride (which is what Oregon uses) to lower the freezing point.

So yeah, probably any chemical has the ability to influence the temperature, you'd have to feeze distilled and pour a fresh cup of distilled water.

Oh yea, thats what i ment, it lowers the freezing point :doh: my excuse is "its late" and I’m trying to quit smoking. so i've been in a little bit of a strange mood lately... Lol

so Martin, why don’t you remove the back of your DMM and take a look inside for some calibration pots, the nicer DMM's have silk screen writing on the PCB of what they correspond to, so hopefully you have something like °c, or Temp written next to one.

[Martinm210]

Oh yea, thats what i ment, it lowers the freezing point :doh: my excuse is "its late" and I’m trying to quit smoking. so i've been in a little strange mood lately... Lol

Hehe, well I'm the one that completely forgot about "Latent Heat". I'm getting old...:D


FYI, altitude and atmospheric pressure does have some influence on the freezing point too, but apparently it's so small it's negligible:

http://www.newton.dep.anl.gov/askasci/env99/env017.htm

[Exahertz]

judging by your results your offset is like 5-6 degrees too high, but the scale looks almost right on the money. at least within an accuracy of &#177; 1-2 degrees

[Martinm210]

Yeah, I popped open the multimeter and there is litterally a lucky number 13 variable resistors in there.

VR1 through VR13....hmmmm

[THRASHER2]

ESU Reptile Digital Thermometer is on Sale right now :clap:

http://www.petco.com/product/100375/ESU-Reptile-Digital-Thermometer.aspx

Internet Price: $9.99
Internet Sale: $6.97
Sale Ends: 1/17/2008

EDIT: here is a 15&#37; off coupon code: PETCOEA

[Martinm210]

ESU Reptile Digital Thermometer is on Sale right now :clap:

http://www.petco.com/product/100375/ESU-Reptile-Digital-Thermometer.aspx

Internet Price: $9.99
Internet Sale: $6.97
Sale Ends: 1/17/2008

Sale ends tomorrow!...hehe

[Exahertz]

hahaha,

ok, now hold up before you start cranking those things....

incase you change something that is set and shouldnt be chaged, you want to measure some known values for voltage and resistance. write them on the thing you tested and keep them around incase you need to calibrate back to those values.

for example, pull out a D cell battery and measure it, write its voltage down,

now measure a resistor from something and write it down.

if you chage something like voltage offset, you can use the battery to set it back. just dont let the battery sit for days or change temperature too much before doing so...

turn things slowly (1/4 to 1/2 turns at a time) and count the turns you make. i like to make reference marks with a marker so i can go the number of turns back to its origional position and then line the marks up.

good luck!

[Martinm210]

hahaha,

ok, now hold up before you start cranking those things....

incase you change something that is set and shouldnt be chaged, you want to measure some known values for voltage and resistance. write them on the thing you tested and keep them around incase you need to calibrate back to those values.

for example, pull out a D cell battery and measure it, write its voltage down,

now measure a resistor from something and write it down.

if you chage something like voltage offset, you can use the battery to set it back. just dont let the battery sit for days or change temperature too much before doing so...

turn things slowly (1/4 to 1/2 turns at a time) and count the turns you make. i like to make reference marks with a marker so i can go the number of turns back to its origional position and then line the marks up.

good luck!

I marked them all first, and I'm a good guesser, only had to tinker with 4 of them before I found it...:D

http://img407.imageshack.us/img407/478/thermometerbummercorrecgj9.jpg

http://img407.imageshack.us/img407/5136/thermometerbummerbetterxn3.jpg

I also noticed you really need to stir it hard to get the temperature down stable, it'll go back up .5 degrees or so if it's not being stirred well.

Not perfect, but much better...:up:

[Exahertz]

:up:
nice!

now all you gotta do is check the scale with some boiling water and your set!

[Martinm210]

:up:
nice!

now all you gotta do is check the scale with some boiling water and your set!

Yeah, I think I'm going to wait until my next thermocouple sensor meter comes in and calibrate them all that way, if nothing else to make them all precise relative to each other. If it's not up to par, I'll just have to buy a good Fluke.

I guess if you also had a good calibrated thermometer like a fluke, you could then use it as your calibration device. For most of the stuff I'm doing, I'm usually more concerned about relative temperatures than accurate temperatures, but wow...I'm suprised this one was off so far. Good thing you can adjust these more expensive types.:up:

[Exahertz]

DM6802B Made by SAMPO... Hmmm, i havnt heard of them, but good price! and K type TC's are the best, IMO!

Persionaly i like EXTECH. i own a 42311 Calibrator Thermometer (http://www.hoskin.qc.ca/uploadpdf/Instrumentation/Extech/hoskin_42311_42312410f9c06afd94.pdf). its nice because it has calibrating features, like temperature output so you can tune other meters (mines uses J Type TCs, which means i cant calibrate my DMMs :( but i might trade it in for a K Type). it works very similar to what you just did but without useing an actual thermocouple or any water. just plug the two meters together and adjust the pots...

[Martinm210]

DM6802B Made by SAMPO... Hmmm, i havnt heard of them, but good price! and K type TC's are the best, IMO!

Persionaly i like EXTECH. i own a 42311 Calibrator Thermometer (http://www.hoskin.qc.ca/uploadpdf/Instrumentation/Extech/hoskin_42311_42312410f9c06afd94.pdf). its nice because it has calibrating features, like temperature output so you can tune other meters (mines uses J Type TCs, which means i cant calibrate my DMMs :( but i might trade it in for a K Type). it works very similar to what you just did but without useing an actual thermocouple or any water. just plug the two meters together and adjust the pots...

Yeah, now that I look over the specs again of the one I ordered, I'm not even sure it has resolution out to .1:( I saw .1C in the accuracy section, and missed the resolution spec.

Oh well. I might have to look for the Extech or a Fluke...:up:

Maybe this:
http://cgi.ebay.com/EXTECH-EA10-EASYVIEW-DUAL-INPUT-DIGITAL-THERMOMETER_W0QQitemZ180206789026QQihZ008QQcategor yZ50974QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

or an older multiple coupler meter like this:
http://cgi.ebay.com/FLUKE-2176A-DIGITAL-THERMOMETER-WITH-10-THERMOCOUPLES_W0QQitemZ290199170839QQihZ019QQcateg oryZ50974QQssPageNameZWDVWQQrdZ1QQcmdZViewItem

[Xilikon]

Putting a probe in the intake and outake of the radiator is pretty useless since you are looking at possibily only 1C delta and the sensors is accurate to within 1C :(

IMHO, just use one to check the water temperature and the best place would be just before the radiator since it will be the hottest point (just 1C but still...).

[Pete]

Martinm210 - You never stop tinkering do ya, good stuff dude

[Xilikon]

mcoffey, I made my comments based on common sense but I'm also curious what the experts really think about that. I'm also wondering where it would be best to place the probe.

[Jedda]

Very easy to forget how fast the water is racing around the system when you start thinking this out.

[Jimbo Mahoney]

Do all DMMs have pots to change the internal calibrations like this, or only the better ones?

I recently bought a new DMM as it had a continuity tester on it, but it also came with a thermocouple. Now, this thing was only £15 (~$30) so I wonder if it has pots for adjustments..

I'll open it up later

Thanks for this thread.

[Martinm210]

Martin,

Where do you think the best place is to put the probes in the loop for monitoring temps? I've kicked around a couple of ideas but haven't decided yet on my bench. I was thinking of placing 2. One directly after the rad, and the other right after the CPU block. I know everything pretty much equalizes throughout the loop, but was wondering if placement skews the deltas.

Thoughts?

andyc

If I had my pick of a dedicated location, it would be right after the rad and before the CPU block because in the end, this is what you are cooling with.

With that said, I've been doing mine post CPU block because that's where my T line is. It's a bit warmer by a degree or so, but like any testing. It's still a good relative result as long as everything is tested the same and enough tests are taken to get a good average.:up:

[SiGfever]

I just use my old Fluke 52, it has been a good meter and serves me well. I need to recheck it soon to see if some tweaking is in order.

[Jimbo Mahoney]

Thanks for this thread!

OK even my cheap £15 / $30 DMM has pots and luckily I found the right one first time.

I can't get it to read correctly at both 100'C (99'C at my pressure and altitude) and 0'C though - it's a couple of degrees high at the low end.

I actually found a cheap digital thermometer I bought from a website was extremely accurate, i.e. better than 1'C, although the website describes it as +/- 1'C. For those of you in the UK, it was here:

http://www.thermometersdirect.co.uk/acatalog/Thermometers_Direct__Waterproof_and_Bath_Thermomet ers_13.html

'Splash Proof T-Shape Stem Blue Thermometer' near the bottom of the page - only £11.75.

[Martinm210]

I havn't tinkered with all the pots a bunch yet, but like Exahertz noted, there could be one adjustement for up and down and one for scale.

If you have both, then scale pot would increase or decrease the spread between the readings at freezing and boiling pt. Havn't tried it myself, but it would make sense.

[Kayin]

Martin, I may have a Fluke that would do what you need, will check after I make this laptop LVDS cable...

Nope, it doesn't have a temp setting. If it did, I'd have been willing to help you out...

Fluke 117, I'll look it up to be sure...

Actually, I stand corrected.

http://us.fluke.com/usen/products/AccessoryDetail.htm?cs_id=8572(FlukeProducts)&catalog_name=FlukeUnitedStates

Yes, it does have a K-type thermocouple available. Badass electrician's meter and thermocouple in one, LMK if you need it...

[Martinm210]

Kayin;2707204']Martin, I may have a Fluke that would do what you need, will check after I make this laptop LVDS cable...

Nope, it doesn't have a temp setting. If it did, I'd have been willing to help you out...

Fluke 117, I'll look it up to be sure...

Actually, I stand corrected.

http://us.fluke.com/usen/products/AccessoryDetail.htm?cs_id=8572(FlukeProducts)&catalog_name=FlukeUnitedStates

Yes, it does have a K-type thermocouple available. Badass electrician's meter and thermocouple in one, LMK if you need it...

Thanks!
I've got my other meter sitting at the post office, so I'll give that one a try tomorrow after picking it up. Hopefully it's accurate and calibrated already, but I'm going to check it.:up:

[Justintoxicated]

Coralife FTL, I noticed the same thing when using those on my fishtanks.

[Jimbo Mahoney]

I havn't tinkered with all the pots a bunch yet, but like Exahertz noted, there could be one adjustement for up and down and one for scale.

If you have both, then scale pot would increase or decrease the spread between the readings at freezing and boiling pt. Havn't tried it myself, but it would make sense.

Well my cheap DMM only has three pots, so I'm guessing one is for V, one is for A (or Ohms) and one is for temp...

[Exahertz]

Well my cheap DMM only has three pots, so I'm guessing one is for V, one is for A (or Ohms) and one is for temp...

well you should have ACVolt, DCVolt DCAmp, ACAmp, Ohms and Temp. but one of those pots should be for temp...

[Jimbo Mahoney]

well you should have ACVolt, DCVolt DCAmp, ACAmp, Ohms and Temp. but one of those pots should be for temp...

I know. I posted earlier that I had found it. I was saying I doubt I have a pot to change the scaling of the temp, only the absolute value.