Let's design and build a better hygrometer, once and for all..!

[Mario Proulx]

Bob, a thought I just had about your device with the probe. As long as they both read the same at ambient temperature, as well as at or near the dewpoint, methinks it could be deemed precise enough. Could you try yesterday's tests, with the glass, using it, please? The inaccuracy(and uncertainty) of last night's test wasn't so much because of a lack of precision, but because with only one good thermometer that I trust, I could only read the dewpoint accurately, and not the air temp at the same moment. If we can read both, simultaneously, the simple glass test should prove plenty precise and repeatable, I suspect. Looking at my photos again, in full resolution, shows the difference between the reading from the moment the dew cleared from the glass, then the moment it reappeared(after introducing more ice) was closer to, and maybe less than, 0.2°C! And again, owing to the fact that the glass was still cooling when I grabbed the camera and shot the second photo, it should be safe to eliminate half of that, leaving us within one tenth of a degree of the true dewpoint. All that is missing is a similarly precise measurement of the air temperature at that same moment, and bingo! A very precise RH reading in under 5 minutes! Also interesting to note was that in the shop, the same glass, topped with beer, was sweaty, and would remain that way until I drank it all, but when I went back into the house, which is a lot drier than the shop, it would clear of all sweat before I drank as much as half of it. Point being, it is a rather precise and quick "tool"....

And while we're at it, is there a film or probe(again, within cost reasons) that could sense the moisture, or lack of, in minute amounts, on the glass or glass plate? that would signal the moment the dewpoint is reached, if cooling the glass, or the moment the dewpoint was passed, if warming the glass.

[Nick Middleton]

The thermocouple uses a 2-conductor cord. At the end that is being used to take the reading you have the insulation stripped for both conductors and the 2 conductors are joined to a single-element (soldered?). The exposed-part of the wires is where the reading is taken from. I believe the principle is that both conductors are dis-similar metals and they create a current (from reaction of the dis-similar metals touching) that will be altered when the temperature around them varies.

I would think the mass of the little element and a tiny bit of glue might be negligible, or there might be some way to factor-out their influence on the numbers.

[Mario Proulx]

Nope, you're right, I think, that the mass of the tiny element should be negligible. The mass I was talking about what that of a completed "probe" type.

Where does one find bare elements, and devices to "read" them?

Obviously, electronics are not my forte...

[Bob Gramann]

I did what I could quickly do today. The glass came out of my freezer and into my workshop. I had to breathe on the glass to see the fog. As you can see, it the sensor didn't couple very well to the glass.

I'm sure the glass was colder than was displayed. The room is at a calibrated 41% humidity. Ignore the electronic readout. The sensor looks as if it is encased in silicone caulk. If I wanted to sacrifice the device, I could dig it out and see if I could get a better attachment to the glass.

As Nick described, a thermocouple is just two dissimilar pieces of metal joined together. It produces a voltage. I think the sensor here might be a thermister which is a bit easier to use in this application--the thermister's resistance varies with the temperature. The advantage of buying a cheap device like mine and tearing it apart is that the electronics are already provided and grossly calibrated.

[Kary Karahadian]

this is the multimeter and thermocouple i use when i bend:

http://www.amazon.com/Extech-EX330-Auto ... B000EX0AE4

[Nick Middleton]

As Kary Mentioned, Some DMM's come with them. I have a Fluke 189 that came with one.

You can also buy pieces, kits, etc. McMaster Carr has some.

http://www.mcmaster.com/#thermocouples/=ki979d

[David King]

What if we started with a steam jet from a pressure cooker/espresso machine and measured how far from the orifice that steam is visible? I imagine atmospheric pressure might affect something and I'm sure all kinds of problems with measuring the steam plume length will pop up but perhaps it's not totally crazy?

[Bob Gramann]

With David's plan, the act of measuring will change the humidity in the room.

[David King]

i.e. crazy <g> Also controlling the pressure in the steam pot would take some effort.

[Steve Senseney]

To add the the confusion--

What is the mechanism in a digital hygrometer? Can it be replicated or improved?

Are we all trying to invent the wheel, when it is already invented?

[Bob Gramann]

I think that the digital hygrometer has a carbon plate exposed to the atmosphere. Moisture in the atmosphere changes the resistance of the plate. A compensation factor for temperature must be applied. Dust on the plate changes everything. All of the plates get dusty. The calibration on the cheap digital ones is almost chance. I don't know how the expensive ones work and why they might work better.

[Mario Proulx]

Steve, my goal is a simple device that is very reliable and precise, enough so that we can use it to calibrate or check our digital and mechanical hygrometers against it. Right now, my trusty old Radio Shack digital unit(an older one from when they actually had adjustable pots in them for spot-on calibration) is pretty much shot, I do believe. I friend just gave me an old Taylor "Mason's Hygrometer", which is a wet/dry bulb deal, and after washing the wick and checking that the two thermometers read the same at various temperatures, it tells me my RH is at 55%, where everything else tells me I'm in the low 40's. Frustrating? Yes, to put it mildly!

So, onward with the ultimate hygrometer project!

BTW, so far, the thin glass, water, ice cubes and thermometer deal is proving to be just about the best thing going. Methinks I'll buy a pair of very precise, matched mercury lab thermometers and call it good. I can even simplify it one full step by keeping the water-filled glass and thermometer in the fridge; when I need it, just pull it out and set it on the bench, grab a chair, and wait for it to clear, while sippin' on a homebrew....

[Bob Hammond]

Hmm, I like Johnny McC's idea. My IR meter's instruction says that it measures a circular spot that is the same diameter as the distance from the surface. Maybe one could glue a dot of black sandpaper, or spray a dot of flat black paint on the face of the mirror. Another thing to keep in mind would be to prevent an ambient light source (esp. incandescent) that might be reflected by the mirror to the IR meter.

[Bob Gramann]

I have a mason's hygrometer. When placed in front of the vacuum hose (as you suggested years ago), I get consistent, repeatable results that match those I get when I use the Acurite unit pictured above in a wet/dry bulb configuration. Rather than do the average as you suggested above, I go for the minimum wet bulb temperature. When it doesn't go any lower, I figure the physics can't lie and use the wet and dry bulb readings at that point. I put the wet/dry bulb temperature to humidity equation into a spreadsheet to do the calculations. I get results with much more precision than I can actually measure. Given the measurement tools we have, the 42% I strive for in my shop can be anywhere from 39 to 45%. I decided that that was close enough. I have two Abbeon hygrometers that I check at least once a year against the psychrometer calibration. They hold their calibration very well. Only once have I ever had to change one after the initial calibration.

That said, I think I have finally solved my dehumidifier problem. I bought one of these: http://www.amazon.com/DAYTON-1UHG2-Dehu ... B001OLVNU0
The displayed humidity has nothing to do with the real humidity but when I set the dial so that the real humidity is controlled at 42% on my good hygrometer, I can leave any cheap dehumidfier plugged into it and set so that it's always on. Now that I can use cheap ones, I don't care that they go bad every two years or so.

[Andy Birko]

But you have to admit, that it doesn't get much simpler than a glass, some water, some ice cubes, and one decently accurate thermometer, now, does it? No batteries, no nuthin'... <bg>

You're assuming that the temperature of the outside of the glass is the same as where the thermometer sits. I don't think that's a good assumption to make if you want a really accurate result.

[Mario Proulx]

Andy, that's why I measured the temperature on the way up, and then on the way back down. It's also why I chose my thinnest glass, and filled it with water, adding thermal mass to it. Since the thermometer also has some thermal mass of its own, it remained in the water the whole time. Glass is also a very poor insulator, so its surface temp should be very, very close to its content's temperature. Read my observations:

I just spent an hour trying various methods we've discussed here, and have some observations to share. First, I found a friend who had an IR gun, so with a fresh battery installed, and I went to work with it. It only reads to a full degree, so it doesn't even have decent resolution, but it had good repeatability. My shop fridge is apparently at 40(°F). When I pull a liquid-filled item from it, it fogs-up immediately. First observation is that while in the fridge, all items showed 40°, yet as soon as they came out and fogged-over(seconds), the readings went to 45-46°. Could the dew itself be right AT the correct temperature, as long as the item on which the dew formed is within a few degrees of the dewpoint? Sure seems like it, right now. Next observation is that a mirror or glass plate can be readily read by the IR gun while it is fogged-up; I repeated this test dozens of times, and also on the back of the mirror, so the results are good. next observation is that the plate glass and mirror warm-up to room temperature WAY too quickly to be of any use. My best, most consistent results were with large jars filled with dark substances(beer yeast cultures, in case you're wondering what all lurks in my shop's fridge.. <bg>).

So, an hour's worth of putzing around with a IR gun and the thin glass/ice/thermometer, here's what I got:

- All jars showed 45 to 46(°C) as the dewpoint. Air temp was, according to the same gun, 70°. That calculates to a RH of 40.68%.

- The glass/water/thermometer gave me a dewpoint of 6(°C) at 19.9(°C)(observed with the same thermometer), for a RH of 40.29%

If consistency and repeatability tells us anything, this means we've nailed it pretty darn close! And, the results correspond perfectly with my mechanical hygrometers(the digital unit is dead...) that were calibrated a few days ago at 44%RH and are reading 40-41% today.

The wet bulb test, using the vac. to pull air across the wick, gave me 44% RH, today.
Bob, I do pull air across the wick of my home built wet bulb unit, but the Taylor's instructions didn't suggest it, so I just observed it as the instructions said. But this time, I grabbed the shopvac and pull air across the wick, and it gave me 44% instead of 55%. Like you, I always use the lowest temperature I can get, but my issue is that I can rarely get the same temp, twice in a row! What's happening, I'm sure, is that I'm drying the wick faster than it can remain wet enough to give a proper reading, so with that in mind, the "trick" here would be to use just enough air to get the right reading, and not so much that we dry the wick, so that brings yet another variable into play. And finally, wet bulb test results can also vary if the atmospheric pressure changes, thus, another variable. Up until now, it was still the best method we had, or knew of, but methinks what we've come-up with here beats it by a large factor.

Oh, and one more observation, and a fun one at that. I noticed that there was some dew that had formed on the lowest inch or so of the shop's patio door glass(it's near zero (°F) outside right now), so I pointed the IR gun right at the dew, with the gun within an inch of it, and it gave me a consistent 45-46(°F), and when I move the gun up just far enough away to be on clear, dew-free glass, it read well into the 50+ ranges. Turns out that all I might need to do is measure the dewpoint right on the shop's windows and doors... <lol> Until spring, that is...

So, unless we come up with something better and cheaper, I'm going to get a pair of good thermometers, and stay with the glass/water/ice as my method. Can't beat the reliability and accuracy of a high-quality lab thermometer, and at $20/40 for a decent pair, I doubt we can match it with any available electronics.

[Jason Rodgers]

Ok Mario, for us dense folks (alright, just me), could you please explain again the process you used with the beer glasses, water, ice cubes, and thermometers? I'm having a hard time putting together that first beer glass post (with the pictures) with this post and making sense of the sequence.

[Mario Proulx]

Take a tall, thin-wall glass, fill it with water, some ice cubes, and place your best thermometer in it. Set it on a table or workbench, and gently stir it with the thermometer once it begins to form a bit of fog or dew at the top, where the ice cubes are. This stirring is to ensure that the water temperature is the same throughout then glass. Once the entire outer surface of the glass has fogged, read the thermometer! Record the reading. Now, scoop-out what remains of the ice cubes, and let the water warm back up. Once you begin to see the surface of the glass clearing, begin gently stirring once again, and read the thermometer when the glass is about 1/3 clear. Record this reading. It should be very close to the first reading; if they aren't identical, average them. That is your dewpoint.

Now go to your computer, and enter the air temperature and dewpoint into the on-line calculator I linked to earlier in this thread, and it will spit-out your RH. That's it!

Now, the reason I suggest we read the thermometer when the glass is 1/3 clear is because the bottom of the glass is many times thicker than the sides, so it will remain cooler than the water; once about 1/3 of the glass has cleared, we should be pretty confident we have a good read.

About the only way I think we could improve this method is with a stir plate, and suspending the thermometer; that way, we could just sit-back and watch.... <bg>

[Murray MacLeod]

Talking about dense folks, include me in ...

If you made a "thermocouple" style humidity gauge (like what you have made for your friends and to which I linked earlier in the thread) ) but made it like 3 feet long, (or longer, what the hell) would that not be an incredibly accurate RH gauge (as long as it was accurately calibrated in the first instance )?

[Kary Karahadian]

how about this for a water vessel. thin walled, very little thickness at the bottom to act as a heat sink. I'm liking this. (sorry for the rotation!)