Testing Hygrometers: My Setup

[Ginseng]

The recent post by PorscheC4S reminded me to post this article. I hope it adds something to the discussion that took place in that thread.

I've been using a mish mash of hygrometers from various vendors over the years. Enough is enough. I finally decided to come into the 21st century and go wireless. Why? because you don't have to keep opening your coolers or cabinets to check the conditions. You can verify the humidity and temperature in all your storage systems as easy as pushing a button at your desk.

Of course, all this technology is useless unless you know how accurate your units are so conducting a periodic (every 6 months) salt test is something I consider absolutely essential. Even if you have the best of technology...trust but verify.

One unit that I've used for several years is a now-discontinued hygro from RadioShack. Small, cheap, reasonably accurate; it is perfect for small containers. It has held steady at a -3 offset the entire time I've owned it. This unit is a bargain at under $20. Uniquely, it uses two button cells in parallel for extra long operation on a set of batteries.

I should mention that a NEGATIVE offset as I define it means the hygro reads LOW. That is, if the salt test target is 75% the RadioShack unit would read 72%. Likewise, a POSITIVE offset means the hygro reads HIGH: 78% when actual is 75%.

About a year and a half ago, I bought two units from Little Havana Cigar Cigars. Primarily this was based on the many positive recommendations that turned up in searches in this sub-forum. Unfortunately, the units I received registered RH that was far from actual. The initial calibrations showed +6 and +7 on the two units. Over the time I've had them, the readings have drifted to +7 and +8, respectively. This should be considered unacceptable and I should have returned the units. This potential drift is one reason why periodic calibrations are so important to give you confidence in the readings from your device.

Most recently, I purchase a three-channel Oregon Scientific wireless weather station from Mark Neff at Cigarmony. His prices on this unit were about the best I could find but what really sealed the deal was the special 2-pack sensor special that you could buy with the weather station (which already came with 1 sensor). This was a stroke of genius as it makes perfect sense that a person buying a 3-channel station might want two more sensors to maximize the potential of the system. Kudos to Mark.

Here is a photo of all my operational hygrometers. The Oregon Scientific base station is at the back. This unit is super cool. It even tells me the time and date by jacking into the signal from the atomic clock in Colorado. It is also smart enough to forecast weather based on the barometric reading and to calculate the moon phase from the built-in calendar. Very smart. Note that for easy reference, I have labeled the base unit and the remote units with the offset and the target reading. These three sensors were pretty darn close to target with one at zero offset, one at -1 and one at +2. I'd consider this acceptable. The smaller unit on the right is the THGR122N sensor that came with the base station and the two larger ones on the left are the new THGR268 sensors from the Cigarmony 2-pack. The new units are identical in operation as far as I can tell but with the added convenience of a larger display and a battery door that does not require a screwdriver to open. (Note: sharp eyes will notice that OS hygrometer #3 is labeled "+2 67" instead of "+2 69" as it should be. I've corrected that on the unit but sure as heck am not going to go drag all the hygros out for another family picture. )

Here is a photo of my hygrometer calibration box. It is a thick-walled, gasketed unit from Lock&Lock. You can find these fantastic boxes at Target. I consider them superior in overall design to the venerable Rubbermaid snap-top containers. Since I have so many hygros, it would have been inefficient to use the small plastic bags that came with a popular one-bag hygrometer calibration kit. In fact, the new THGR268 sensors just fit into the bags, but it's too tight for my comfort. I bought two calibration kits just to check my salt setup and ended up using both the calibration pillows along with a dish of salt solution of my own. I also checked this calibration at two temperatures, 60F and 70F and across that range, the measured relative humidity did not change at all for any of the hygrometers. This result gives me the confidence that within the range of temperatures that my storage containers experience, the readings will be accurate and reliable.

Here is a close-up of my salt slurry dish. Keep in mind that the science says that 75% is achieved by using a saturated solution. In my experience, a wet slush like this gives more reliable results than a slurry that appears dry. I like there to be a layer of liquid at the top of the dish.

And that's pretty much it. I am really loving the OS weather station and the ease and peace of mind it provides. I suspect it won't be long before I spring for another base station and additional sensors. This thing is just too damned good not to have watching over my cigar collection.

PS. I use self-conditioned silica crystal kitty litter and/or propylene glycol water solution to manage the humidity in my coolers.

PPS. Whenever I changed conditions in the calibration box (added hygros, moved from higher to lower temperature conditions) I gave the system 24 hours to equilibrate. Typically the readings would adjust and stabilize quite quickly over the course of an hour or two but I wanted to make sure the flatline-unchanging readings really were stable. You don't really want to rush things but this suggests that an impermeable container (such as tomato sauce jars I've used in the past), ones with tight seals and/or thick walls will allow conditions to stabilize quickly and well. This is what you want when assessing an equilibrium condition. Single layer ziplocs are not, in my opinion, optimal.

****************************

Here's a little bit more information.

I recently received the replacement adjustable digital hygro from Mark and calibrated in the Boveda kit to read 75% in the bag. I then placed it in a loose NC sticks coolidor along with the +7 offest unit from Little Havana. Now keep in mind that in the bag, the LH reads 82% as it did before and the Adj read 75% as expected. The difference between these two units is 7%. I just reconfirmed that this is the case. All is good, no? No.

When I placed both of them in the loose coolidor right next to each other, the LH read 74% suggesting an actual condition of 67%. The Adj, however, read 70% for an actual of 70%. The difference between these two units is now 4%. This is a big problem. Big. This means that as actual humidity changes, the offset between the "true" and "indicated" humidity for one or both hygrometers changes. In other words, it is impossible to have any confidence in the hygrometer reading except at the exact humidity at which the hygrometer was calibrated.

In the following graph, I plotted the true and indicated values for both hygrometers at two different humidity conditions. For the purposes of this chart, I assumed the Cigarmony was giving me the correct reading in both cases but it does not matter which one I choose as the reference. I am also assuming that the Boveda kit is giving me 75% as advertised. I also gather that my coolidor is at a "different" humidity although now I am not sure exactly what. For purposes of comparison, I then regard the Cigarmony unit as "correct" at both conditions. This then provides the baseline against which another hygrometer can be compared.

What would be ideal would be to calibrate a hygrometer at two different, known humidities. This would give you the slope of the line of "indicated" against "true" humidity which should be 1.0. In the above case, the slope of the LittleHavana unit is greater than 1.0 which is not good.

The essential question that falls out of this observation is this: Which hygrometer, if any, is giving the proper reading across different humidities? One or both of these two hygros is off but we do not know which and by how much. In essence, what the 4% and 7% differences means is that one or both hygros have a response curve that is faulty. To make a determination about which one it is, I would need to put both in another calibrated environment, for example a Boveda 65% pack and check indicated readings in that environment. That would seal the deal although three points would even be better, say 65%, 70%, and 75%.

In an attempt to make things just a bit clearer, I've fabricated "Ideal" and "Worst Case" calibration curves for four hypothetical hygrometers.

Ideal or desirable case: All parallel curves with very small spread at either end. Small, consistent offset.

Worst or undesirable case: Each curve has a wildly different slope and offset.

I hope that you will find this illuminating is some sense. Although I certainly obsess about the phenomena, my actual maintenance routine is quite simple otherwise I'd rapidly come to hate this little hobby.

Wilkey

[El Presidente]

Wilkey...if you were female..... I would marry you